FindSpots User Guide - The Open Microscopy Environment
Contents
1. Welcome David Schiffmann No recent project create new Most recent dataset UsIRNA201005 Popup Update Clear Selected Clear All Task Status N steps Start time Elapsed time PID Executing Find and track Oo spots FINISHED 2005 08 30 11 29 18 00 10 22 25670 Executing Find and track spots Executing Image server stats FINISHED FINISHED Importing images FINISHED Imported 7 images from 7 files 7 scanned 0 unknown format 0 duplicates 0 errors 285 2005 08 30 16 19 24 2005 09 02 13 09 08 00 13 12 00 22 47 25667 Executing Find and track spots FINISHED 2005 09 02 14 55 42 01 53 18 Executing Find and track spots Executing Find and track spots Executing Find and track FINISHED spots FINISHED FINISHED 2005 09 05 17 54 20 2005 09 06 13 53 23 00 06 05 00 05 15 Executing Image server stats FINISHED 2005 09 06 14 01 52 00 00 35 Executing Find and track spots Executing Image server stats FINISHED FINISHED Importing images FINISHED Imported 1 images from 1 files 1 scanned 0 unknown format O duplicates O errors 2005 09 06 14 01 04 2005 09 28 14 20 17 00 01 23 00 00 26 Importing images FINISHED Imported 2 images from 2 files 2 scanned 0 unknown format 0 duplicates 0 errors 2005 09 28 14 18 58 00 01 42. this can be done by editing the name and clicking Save unless the dataset has become locked by OME in which case one should contact the local OME database administrator or the OME developers Selecting files to import NB Be careful to select the right images to import into the chosen dataset as it is not readily possible to delete images once imported though they can be removed from datasets and or reassigned to different datasets later on One can add images to datasets later on using OME s Add Images and Remove Images commands to add or remove already imported images into that dataset see below To find the files to import use the Directory Listing window in the lower portion of the page Click on directory names to navigate to the directory where images for import are stored Once a list of images is displayed choose the files to import either click on select all to select all files in the current directory or check the boxes next to the files to be imported Once the files have been selected click on Add to queue which will cause these files to be added to the queue on the right hand panel titled Import Queue Files can be removed from this list prior to starting the import by selecting the check boxes next to them and clicking Remove from Queue Files can be added to the import queue from different parts of the directory structure Once you are satisfied that the correct files have been chosen for
3. 11 01 14 02 14 iA 11 r3d_D3D dv David Schiffmann OME 2005 11 01 13 36 46 ModuleExecution s ModuleExecution s 2 iA 13 i r3d_D3D dv David Schiffmann OME 2005 11 01 13 26 20 iA 13 ii r3d_D3D dv David Schiffmann OME 2005 11 01 13 38 14 ModuleExecution s ModuleExecution s iA 15 r3d_D3D dv David Schiffmann OME 2005 11 01 13 50 13 iA 16 r3d_D3D dv David Schiffmann OME 2005 11 01 13 57 27 iA 09 r3d_D3D dv David Schiffmann OME 2005 11 01 13 59 46 ModuleExecution s iA 12 r3d_D3D dv David Schiffmann OME 2005 11 01 13 37 20 ModuleExecution s iA 14 r3d_D3D dv David Schiffmann OME 2005 11 01 14 03 11 ModuleExecution s iA 17 r3d_D3D dv David Schiffmann OME 2005 11 01 13 42 33 The view that appears after clicking the More info link for a dataset Welcome David Schiffmann o recent project create new ent dataset UsIRNA201005 Popup Find the image you wish to open it may be necessary to move between pages using the blue gt signs at the top of the page and click on the image title blue underlined name A view like the one below will appear This is the Image Detail view for the selected image This view can be reached from any thumbnail displayed on any OME page by clicking the upper left corner of the thumbnail Clicking on any other portion of the thumbnail will launch the OME image viewer It is not necessary to click More
4. 47 8 36 9 17 7 34 30 47 8 36 9 17 7 535 8 18235 536 3 48 36 8 17 6 30 27630 1437 0 992 38 3 37 18 3 58 49 38 3 37 18 3 457 8 26646 459 4 38 2 36 7 18 1 31 27631 1437 o 992 111 1 44 18 1 34 31 111 1 44 18 2 345 8 11772 346 2 111 2 44 2 18 1 32 27632 1437 i 992 69 51 2 24 9 378 273 69 1 51 2 24 9 589 6 223518 591 3 69 8 50 5 23 7 33 27633 1437 o 992 110 8 67 6 19 2 33 33 110 9 67 6 19 2 466 1 15458 468 4 110 7 67 5 18 9 34 27634 1437 0 992 8 2 68 4 22 3 47 47 8 2 63 4 22 1 156 6 7385 157 1 8 2 68 4 22 3 35 27635 1437 o 992 20 77 7 19 32 29 20 1 77 6 18 9 169 5 5439 170 20 1 77 6 19 36 27636 1437 0 992 44 4 42 1 24 1 51 43 44 5 42 2 24 500 6 25601 502 44 3 41 9 24 2 37 27637 1437 oO 992 128 9 54 8 23 4 33 31 128 9 54 3 23 5 406 5 13511 409 4 128 8 54 9 23 3 38 27638 1437 o 992 70 59 7 24 1 53 44 70 59 7 24 2 577 6 30696 579 2 69 8 59 7 24 39 27639 1437 o 992 38 3 48 2 25 56 45 38 3 48 2 25 457 4 25641 457 9 38 3 47 8 24 8 40 27640 1437 o 992 92 6 641 24 8 43 37 92 6 64 1 24 7 548 9 23626 549 4 92 8 64 1 24 6 41 27641 1437 o 992 41 33 4 25 8 34 32 41 33 4 25 8 441 7 15041 442 4 41 3 33 3 25 6 42 27642 1437 0 992 73 5 38 5 25 4 31 27 73 5 38 5 25 4 561 1 17433 562 4 73 5 38 7 25 3 43 27644 1437 o 992 95 9 54 7 25 6 37 32 95 9 54 7 25 6 590 7 21873 591 2 95 8 54 8 25 3 44 27646 1437 o 992 112 6 SS 27 4 62 55 112 6 55 27 6 485 2 30192 487 112 7 55 1 27 4 45 27647 1437 o 992 17 57 24 9 32 29 17 57 25 265 3 8518 266 2 17 57 24 9 46 27650 1437 0 992 113 5 76 2
5. Enable Macros to continue otherwise the spreadsheet will not function as needed 8 Note that because of the need to use Microsoft Excel s programming language this is only possible using Microsoft Excel and will not work with other spreadsheet programs Any of the Download as text links found in table views in OME pages will download a tab separated text table to any spreadsheet viewer capable of reading tab separated text tables 24 Using the spreadsheet connecting to an OME server After opening the OME Excel spreadsheet select the worksheet entitled OME connection This worksheet will appear when the OME Excel file is opened Here one needs to enter the OME server name and userID and then log in with the password when prompted This provides the spreadsheet with the current OME session key and allows it to transfer data from the OME server this must be done prior to other analysis steps If one is running the OME server on the same computer as one is using to run this Excel spreadsheet then server name is simply localhost The OME Username should be the same one used when installing OME or any other valid OME user It is convenient to save the spreadsheet at this point in order to retain the server and usernmae settings for future use Once these values are entered into the worksheet click the Login to OME button to provide a password and retrieve a SessionKey from the OME server The SessionKey wi
6. Info in the dataset detail page if the desired image can be recognized from it thumbnail 17 OME Home gt gt Image Detail Project Dataset Open Microscopy Environment 4 Welcome David Schiffmann cent project create new UsiRNA201005 Popup Image iA 08 r3d_D3D dv Sikar kA Id 624 Owner David Schiffmann Group OME Created 2005 11 01 13 48 44 Imported 2005 11 01 13 48 44 Original File iA 08 r3d_D3D dv View Image in VisBio Description save Projects Datasets Images Module Executions Chain Executions Other Create a Template Annotate Images Search by Annotation Import Spreadsheet Your Current Textual Annotation Save Mark Invalid View all O Annotations Import Export Image s Find Spots Import Modules Create a custom annotation of _ Select a Semantic Type Execute Chain View Chain Results Datasets More info Define a protocol UsiRNA201005 Tasks Logout Pixel Sets More info Pixels info Module Executions More info Image import ran on 2005 11 01 13 48 44 Stack statistics image server ran on 2005 11 01 14 55 49 Track spots ran on 2005 11 01 14 52 53 Find spots ran on 2005 11 01 14 46 14 Stack statistics image server ran on 2005 11 01 14 40 50 Plane statistics image server ran on 2005 11 01 14 19 27 Stack statistics image server ran on 2005 11 01 14 08 24
7. OME Project Dataset F amp kd Projects Datasets Images Module Executions Chain Executions g z e 8 Create a Template Annotate Images Search by Annotation Import Spreadsheet Import Export Image s Find Spots Import Modules Execute Chain View Chain Results Define a protocol Tasks Logout Internet zone Home gt Analysis gt Image Search Look for image From Dataset UsiRNA201005 X C That match criteria Name av Description AY Created AY Inserted Av Owner AY David Schiffmann Group A V All Id av Display Results as Summaries Names Open Microscopy Environment lof2 gt iA 01 r3d_D3D dv David Schiffmann OME 2005 11 01 13 30 41 Results iA 02 r3d_D3D dv David Schiffmann OME 2005 11 01 13 34 25 ModuleExecution s iA 04 r3d_D3D dv David Schiffmann OME 2005 11 01 13 35 09 ModuleExecution s iA 07 r3d_D3D dv David Schiffmann OME 2005 11 01 13 58 56 ModuleExecution s iA 05 r3d_D3D dv David Schiffmann OME 2005 11 01 13 30 00 iA 03 r3d_D3D dv David Schiffmann OME 2005 11 01 14 06 10 ModuleExecution s iA 06 r3d_D3D dv David Schiffmann OME 2005 11 01 13 45 40 _ModuleExecution s ModuleExecution s 3 iA 08 r3d_D3D dv David Schiffmann OME 2005 11 01 13 48 44 ModuleExecution s ModuleExecution s 2 iA 10 r3d_D3D dv David Schiffmann OME 2005
8. formulae can be generated and automatically applied to all worksheets and is a convenient place to generate a summary table from all the images in the dataset for statistical analysis or plotting Before preparing formulas it may be helpful to make a separate list of the dataset this is from and what the different emission wavelengths correspond to e g what antigen is labelled in each channel and the channel index number 0 1 2 etc Then use this to prepare formulas as required Any formulas that need to refer to the worksheet should be generated using an arbitrary worksheet filename and placed in one cell as an example Clicking on the Fill down formulas button will then generate versions of the formula corresponding to each of the worksheet name forming a column with a suitably modified formula in each cell Repeat this to generate a new set of formulas in another column as needed This should be done before any spaces are inserted between rows otherwise the automatic formula generation will stop when it reaches one of these spaces The way in which the data is analysed next depends on the needs of the user For example for any one image it is possible to generate scatterplots of volume vs mean intensity or x position vs total spot intensity or to sum the total intensity of all spots in the image in each of the different wavelengths Post import checks After importing data into Excel and preparing the formulas and a summ
9. heading will allow you to choose a dataset to process To do this first click on dataset under the Search menu and you will be presented with a view as shown below containing a few images from each dataset eoe Search for Dataset e ff 2 SA Back Refresh Home AutoFill Print Mail o http bonedancer openmicroscopy org uk perI2 serve pl Page OME Web Search amp SearchType 0ME Dataset e Page Apple Apple Support Apple Store Mac MacOsx Microsoft MacTopia Office for Macintosh MSN Welcome David Schiffmann Open Microscopy Environment z4 No recent project create new Most recent dataset UsIRNAnoc161005 Popup OME Home gt Analysis gt Dataset Project Search Results Dataset Other 06 07 04 05 140405 58 image 10 12 04 335 and 341 Search ei Datasets Images That match criteria Module Executions Description AY sara me 11 10 05 UsiRNAmon111005 21 images 15 04 05 monastrol 335 s Other Create a Template Annotate Images Id Av Lea oe ne 21 04 05 Bub1 floxed a s 21 04 05 BubR1 floxed Import Spreadsheet L Mages Import Display Results as Export Image s Summaries ETT O Names Find Spots UsiRNA201005 52 imag Import Modules Execute Chain View Chain Results Define a protocol Tasks Logout UsiRNAtax131005 31 ima age Dg 28 03 l 1a and 31 imag exp 29 07 04 335 341 32 grO 2 images movi
10. is a binary image Keep adjusting the sliders so that the binary image displays only the spots you want found The numerical value of the white black level is a threshold to use for FindSpots You can use this numeric value directly when applying the Absolute threshold method in FindSpots Usually you will have a sprinkling of pixels that are not in your spot but are still above the threshold noise in other words Since these are normally single pixels or generally much smaller than the spots you re interested in the minimum spot volume parameter in FindSpots lets you filter these out the spot volume is the number of pixels in the spot 13 Moving from an absolute threshold to a statistical threshold As mentioned above the absolute threshold method is usually not as useful as the statistical methods which automatically calculate the threshold anew for each image This is because the absolute pixel intensity reported for the white black level is not very robust i e it needs to be varied to account for differences between images which may be due to differences in background between images or perhaps photobleaching in different timepoints A more robust value that is less sensitive to differences in background is therefore required For this reason you can specify the threshold as some number of standard deviations sigmas above the mean or geometric mean In the web based viewer described above click on the Stats but
11. with similarly named sub directories as for the before cropping sub directories Cropped edited files will be saved in here This has a number of advantages First it allows one to go back to the original file if one needs to adjust the cropping Second it is often helpful to look back at the original files to see the full context of a closely cropped image e g to help confirm the cell cycle stage Crop the images as needed to produce images that do not have too much non cell black space around the cells and which avoid including kinetochores from neighbouring overlying cells Because software is often limited to rectangular shapes for the cropping it may be necessary to compromise and miss out some kinetochores from the desired cell in order to avoid simultaneously including kinetochores from neighbouring cells NB it is best to scan through all the Z sections of each image because a file can appear to contain only one cell but the other cell s may only be visible in a sub set of the Z sections i Note that it is okay to collect different numbers of slices in different images or to vary the XY size of the images collected and images will likely all be cropped somewhat anyway Ideally the Z sections should extend above and below the region of interest here the kinetochore containing region by about the same amount for each image since this could skew the overall mean pixel value which may be used to determine an automat
12. 234761 2115 1 136 2 72 4 7 389 4 50714 456 9 136 1 72 1 5 1 946 1 112235 1011 1 136 2 72 1 5 2 1186 7 132526 1193 9 79 4 81 7 8 1408 8 221584 1691 5 79 4 81 7 7 8 304 4 47307 361 1 79 5 81 9 8 3 900 124096 947 3 79 3 82 1 7 9 1184 8 157497 1202 3 88 4 41 2 11 3 1570 5 310750 1806 7 88 4 41 1 11 2 323 2 66919 389 1 88 8 41 1 11 8 892 1 160919 935 6 88 6 41 2 11 7 1157 8 201844 1173 5 121 2 36 9 9 1740 6 88963 2169 8 120 9 36 8 9 285 5 13610 332 121 36 8 9 6 893 5 38345 935 2 121 36 8 9 6 1273 5 52308 1275 8 87 3 56 1 1330 2 214925 1568 8 87 3 56 1 311 3 48874 356 7 87 2 56 3 868 5 124748 910 6 87 4 56 4 945 5 131302 958 4 66 19 9 1593 2 201766 2058 8 66 19 8 303 37141 379 19 9 898 92847 947 4 66 19 8 902 6 90018 918 6 9 5 1803 9 326411 2133 4 lofiS gt gt gt Inputs Numerical output of FindSpots accessed via the web interface this is the same page as in the previous figure but after scrolling down This page shows the image used as the input to FindSpots and all the outputs it produced in a number of tables For example the signal intensities spot integrals are in the Signals output Within the web browser it is usually only possible to view a small portion of each table at a time and it is often easier to view the data after downloading it and opening it with a spreadsheet prog
13. 25 5 41 36 113 5 76 2 25 4 405 7 16720 407 8 113 6 76 1 25 3 47 27651 1437 o 992 55 2 33 9 29 77 69 55 3 34 29 2 486 1 37689 489 5 55 1 34 1 28 8 48 27653 1437 o 992 66 9 698 28 8 38 37 67 69 7 28 8 361 1 13752 361 9 67 69 7 28 9 49 27655 1437 0 992 54 1 42 1 27 4 32 28 54 1 42 1 27 5 555 6 17820 556 9 54 2 42 2 27 8 50 27657 1437 0 992 21 8 83 7 30 5 79 70 21 8 83 6 30 6 151 6 12006 152 21 8 83 7 30 5 51 27659 1437 oO 992 46 24 8 28 34 31 46 24 8 28 316 8 10787 317 3 45 9 25 28 52 27661 1437 oO 992 74 4 87 6 28 1 33 35 74 4 87 6 28 1 216 4 8240 216 8 74 4 87 6 28 1 53 27663 1437 o 992 41 5 37 8 29 3 55 46 41 5 37 8 29 3 416 6 22944 417 2 41 5 38 1 29 3 54 27665 1437 o 992 52 6 46 4 30 6 42 38 52 6 46 4 30 6 541 3 22786 542 5 52 7 46 3 30 3 55 27667 1437 0 992 57 9 62 3 31 6 70 60 57 9 62 3 31 6 490 5 34471 492 4 58 2 62 1 31 5 56 27668 1437 oO 992 117 9 68 30 8 50 42 117 9 67 9 30 8 556 9 27874 557 5 118 1 67 8 30 6 57 27670 1437 0 992 120 3 58 1 32 5 41 40 120 3 58 2 32 4 484 2 19898 485 3 120 2 58 4 32 3 58 27672 1437 oO 992 74 2 57 2 34 5 93 7 74 2 57 1 34 4 561 3 52347 562 9 74 4 56 8 34 3 59 27674 1437 i 992 93 5 71 32 9 30 27 93 5 70 9 32 9 452 1 13656 455 2 93 6 70 8 32 9 60 27676 1437 0 992 89 4 62 34 3 37 32 89 4 62 34 3 589 7 21852 590 6 89 6 61 9 34 4 61 27678 1437 0 992 127 4 67 35 6 52 45 127 4 67 35 6 505 8 26345 506 6 127 5 67 2 35 4 4 62 27680 1437 i 992 114 65 5 39 3 76 69 114 65 5 39 2 503 38307 504 114 65 7 39 6 X i4 4 gt gt i
14. 5 Executing Image server stats FINISHED 2005 10 11 15 41 43 00 06 57 Importing images FINISHED Executing Find and track FINISHED spots Imported 37 images from 37 files 37 scanned 0 unknown format 0 duplicates 0 errors 2005 10 11 15 27 24 2005 10 11 15 50 59 00 21 16 00 10 42 Executing Find and track 0 FINISHED spots 2005 10 11 15 51 26 00 10 06 Executing Image server stats FINISHED 2005 10 14 15 25 19 00 12 49 Importing images FINISHED Imported 35 images from 35 files 35 scanned 0 unknown format 0 duplicates 0 errors 2005 10 14 14 59 45 00 38 23 m Executing Find and track monks FINISHED 2005 10 14 15 42 58 00 13 12 Executing Find and track a FINISHED 2005 10 14 15 44 30 00 11 11 Internet zone Pe Tasks window At any time one can see a list of current or previous tasks in OME by clicking on Tasks under the Options menu 13 Inspection of FindSpots results Visual inspection of overlays After running the analysis it is recommended to check some images to make sure the analysis has worked as desired If the relevant dataset is not currently visible click on Datasets under Search and then click on More info next to the chosen dataset A new window will appear 16 Address
15. 9 In that case using a ThresholdValue of 3 5 will yield a threshold of 398 3 5 x 189 1059 5 Set the sliders to as close to 1059 5 as possible and scan through the optical sections in the image to ensure that a satisfactory thresholding is achieved Repeat this for some other images as desired 12 Starting a FindSpots run Before running FindSpots the dataset on which to run the analysis should be selected To do this click on Datasets under the Search menu and click on the required dataset from amongst the list of datasets This will open a page showing thumbnails small pictures of each image in the dataset Next run FindSpots click on Find Spots under the Analysis menu Enter the threshold and minimum spot volume set as discussed above 14 It is also necessary to specify which emission wavelength will be used to define the spots Note that FindSpots will use the spots defined by this one emission wavelength to calculate the spot intensity within those spots not only for the wavelength chosen but for all the other wavelengths in the image as well This means that one can define spots using a high contrast label and then work out the total intensity of some other channel s within those spots even if those other channel s do not have such a high signal to noise ratio i e even if one would not easily be able to define the spots using those signals alone After checking that the parameters have been entered correc
16. 96 FINISHED 29 9 05 17 45 1436 38 1 95 FINISHED 29 9 05 17 46 1438 39 40 a 7 i44 gt gt i Get Chain MEXes Get Image MEXes SampleQueries OMEconi 9 lt gt ih Ready OSCR OCAPS NUM Importing FindSpots results into Microsoft Excel How to specify the desired dataset and FindSpots run to import data from 2 One will now see a window like this The simplest next step is to click on Get All results to begin the import of numerical FindSpots results spot size total intensity etc for all images in the dataset One can then enter formulae in columns F and or those to the right and click Fill down formulas to automatically generate the same formula for each image in the spreadsheet e g for each image in that FindSpots run but modified to refer to each particular image worksheet in turn Layout of imported data When import of data has finished look through some of the worksheets to familiarise yourself with the layout of the data 28 00e M CenpBsiRNA270905 528 thres xls p D E F G H I J K L M N o 1437 Name iAPCB10_R3D_D3D dv 10 35 Channels Index LogicalChannel _ Emiss Fluor 36539 96591 96593 lt m le execution TheT Threshold Thex TheY TheZ Volume SurfaceArea CentroidX CentroidY CentroidZ GeometricMean Integral Mean CentroidX CentroidY Centroid2 Geor 10 Feature Modul 27605 1437 o 992 10 4 54 4 7 2 39 37 10 4 54 4 7 3 190 7 7450 191 10 4 54 4 7 1 12 27607 1437 o
17. 992 12 77 5 3 8 31 31 12 77 5 3 8 161 4996 161 2 12 77 5 3 8 13 27608 1437 o 992 88 7 348 51 36 31 88 6 34 8 5 2 415 6 15094 419 3 88 6 35 5 1 14 27610 1437 o 992 85 4 48 7 6 2 44 37 85 4 48 7 6 2 442 5 19528 443 3 85 3 48 6 6 2 15 27612 1437 o 992 90 8 40 2 10 8 32 31 90 6 40 2 11 543 7 17504 547 90 4 40 11 4 16 27614 1437 o 992 58 3 43 3 9 35 30 58 3 43 3 9 532 5 18668 533 4 58 2 43 8 9 17 27617 1437 o 992 70 3 51 1 11 7 49 40 70 3 51 1 11 6 501 7 24724 504 6 70 2 50 9 11 4 18 27618 1437 o 992 26 1 100 9 11 7 42 41 26 1 100 9 11 7 137 6 5786 137 8 26 1 100 9 11 6 19 27619 1437 o 992 74 6 37 4 12 5 34 28 74 6 37 5 12 5 452 7 15423 453 6 74 5 37 6 12 6 20 27620 1437 oO 992 48 2 43 9 12 5 32 28 48 2 43 9 12 5 536 3 17173 536 7 48 4 43 7 12 4 21 27621 1437 o 992 23 8 57 5 16 5 51 50 23 8 57 4 16 6 253 9 12976 254 4 23 8 57 4 16 6 22 27622 1437 oO 992 52 1 32 5 13 9 33 30 52 1 32 5 13 9 508 9 16810 509 4 52 1 32 7 14 3 23 27623 1437 oO 992 68 44 5 14 5 34 32 68 44 5 14 5 545 2 18558 545 8 68 44 4 14 4 24 27624 1437 oO 992 111 5 61 14 2 42 37 111 6 61 14 3 541 2 22763 542 111 6 60 8 14 1 25 27625 1437 o 992 74 9 66 9 13 5 30 27 75 66 9 13 6 347 4 10472 349 1 74 9 66 9 13 5 26 27626 1437 o 992 117 3 51 15 2 41 35 117 3 51 15 2 442 18165 443 117 2 51 2 15 1 27 27627 1437 0 992 119 7 64 2 15 9 44 41 119 6 64 2 16 1 500 9 22149 503 4 119 8 64 3 15 7 28 27628 1437 oO 992 113 3 73 16 9 39 33 113 3 73 16 9 385 4 15081 386 7 113 3 72 8 16 3 29 27629 1437 o 992
18. Execu Module Execution poe fo Stack means Secr ah d seack geomeans fEmpseatine __ stack stomas Stack minima Evpseatime 09 __ srack manny feapseaTime wil stack geosiomas Elapsed Time T 13 33294 pe o oi B satus FINISHED ferormesese id View graphic overlay 7 S S Dataset Other g7 z 3 89 a8 Images Module Executions Chain Executions Other Merge Outputs Create a Template Annotate Images Search by Annotation Import Spreadsheet Import Export Image s Find Spots Import Modules Execute Chain View Chain Results Outputs Timepoint Search within these outputs Download as txt 1 of 4 gt gt gt TheT 0 Define a protocol Feature 666964 666969 666974 666979 666984 666989 666994 666999 667004 667009 667014 667019 667024 ojojojejejojejejefefefe Numerical output of FindSpots accessed via the web interface To view a graphical representation of the results the spots found overlaid over the original image click on View graphic overlay 21 Signals Search within these outputs Download as txt 1 of 15 gt gt gt Background CentroidX CentroidY CentroidZ GeometricMean GeometricSigma Integral Mean Sigma TheC Feature 136 72 2 4 7 1850 7
19. FindSpots OME v2 5 1 user guide v Z slice 31 55 lD lt p Timepoint 1 1 lt i gt lt p Channel Black White i i Level Level 685 On 159 Image Info Stats Save settings SaveasTIFF Preload planes Show all Z Show all T User guide by David A Schiffmann Paul L Appleton amp Ilya G Goldberg Email igg nih gov 24th February 2006 The small print Whilst every effort has been taken to ensure the accuracy of the information in these pages it is subject to change from time to time The National Institutes of Health and the University of Dundee make no representations or warranties either express or implied as to the accuracy of the information in this document or its fitness for any purpose whatsoever In no event will the National Institutes of Health and the University of Dundee be liable for any direct indirect special incidental or consequential damages arising out of the use of the information held in this document The National Institutes of Health and the University of Dundee are not responsible for the content of any external websites which may be referred to here or linked to from the Open Microscopy Environment website Mention of third party software in this document does not imply any kind of endorsement or recommendation The Open Microscopy Environment software is provided as is and any expressed or implied warranties including but not limited to the implied warranties of me
20. Instrument settings none Features 1 of 34 gt gt gt More info Trajectory 89 Trajectory 88 Trajectory 87 Trajectory 86 Trajectory 84 Information about an individual image file This provides access to previous Find spots runs according to the date time they were run To look up the FindSpots output for this image on the right under More info will be a number of headings choose the one called Find Spots not the one called Track spots Look for the one nearest the top which will correspond to the most recent analysis one can always confirm this by the date time details provided therein and by checking that the parameters used for that run such as ThresholdValue MinimumSpotVolume or the thresholding algorithm are as expected Click on View graphic overlay this will display the original image To turn on the display of spots found click on Overlay By default this will only show spots whose centroids centers of mass weighted by intensity of all pixels in the spot are in the current optical section However it may be helpful to simultaneously display spots in all optical sections This can be done by turning on the all Z option click in the circle next to it and a white circle will appear to indicate this option is active one can return to a display of only spots in the current optical section by clicking Show all Z again A similar option Show all T allows the user to t
21. _FindSpots1437_iAPCB10_R3D_D3D d _ Get Chain MEXes _ Get Image MEXes A gt a Ready OSCRL O S NUM Example of layout of imported data A separate worksheet is generated for each image listing the FindSpots results as shown here At the top left of the image red box is a summary of the filename and the emission wavelengths Each row thick blue arrow corresponds to one spot The data is divided into blocks of columns as shown by the red arrows the thick red arrow corresponds to basic spot data as defined by the thresholding whereas the thin red arrows show data specific to each wavelength in turn in the order shown in the box Hence A corresponds to 457 nm fluorescence emission B to 528 nm etc For each image a separate worksheet is generated In each such worksheet there is a header with information common to all the spots in that image and actual spot data laid out in columns one column per channel index number emission wavelength in the same order as given in each worksheet as shown in the screen shot above For example in the above worksheet the left most block of data corresponds to channel 0 457 nm emission the next block to the right corresponds to channel 528 nm and the right most block to channel 2 617 nm Only the first block is shown for reasons of space 29 Processing the imported data To further process the Excel spreadsheet thus generated go to the Get Chain MEXes worksheet this is where
22. ary table it is a good idea to go through this table and look for any potential problems It may be helpful to generate a column giving the total number of spots per image to help detect images with an unusually large or low number of spots which may indicate a problem with the image Other problems could include the presence of images that should not be there as judged by their filename Move any rows each corresponding to one image to be ignored to another part of the spreadsheet or delete them It may be the case that in some images certain spots could not be excluded by cropping due to the restriction to a rectangular cropping shape or otherwise and these will need to be manually removed within FindSpots WebUI find the unwanted spots ID and then delete these from the data table in Excel however it is far easier to simply avoid using such images or crop them suitably before analysis 16 References I T Young J J Gerbrands and L J van Vliet Fundamentals of image processing in V K Madisetti D B Williams eds The Digital Signal Processing Handbook CRC Press Boca Raton FL 1998 51 01 51 81 30
23. at can be saved in a suitable file format Indeed FindSpots amp OME are not limited to microscope images and can also be used to store manage and analyze images from other types of equipment 3 At the microscope image collection advice When collecting images it is important to consider how the image will need to be processed e g by cropping and how it will be analysed as taking this into account at the image collection stage can simplify further analysis Ensure that all images from a single experiment to be compared directly against each other which typically will subsequently be imported into a single or perhaps more than one OME dataset explained later on are collected using identical exposure times per channel i e for any one channel the exposure time is the same for all images and identical binning settings on the microscope where this is 1 an option NB When capturing images of cultured cells if possible ensure that cells are not too closely packed next to each other or on top of each other this will make subsequent image cropping simpler especially if the software package used for cropping only allows one to specify rectangular regions to crop 4 Pre processing preparing images for analysis in FindSpots Images may need to be cropped before analysis i e some regions of the image may need to be removed through all optical sections e g to remove unwanted neighbouring cells and some optical
24. blue circles and the ID number will update automatically An X mark will always be displayed in the blue circle whose ID number is currently displayed to ensure that the spot intended by the user is the one currently displayed e098 341ABRO19_R3D_D3D dv Q Greyscale Map Z slice 24 46 lt D lt p gt Timepoint 1 1 lt gt lt gt Channel rs Each spot has a unique number associated with it Note if show all Z is active showing spots found in ALL optical sections at once then it is possible that the blue circle clicked on may in fact not be from a circle whose centroid is in the current optical section In such a case the viewer will automatically change to a different optical section to display the one in which the spot s centroid lies 15 Importing data into Microsoft Excel using macros Introduction Although it is possible to view and download the numerical data from the various output tables as 7 The feature ID number here should be ignored as it is an internal reference number that provides a unique reference to that feature spot within the entire OME database but it cannot readily be cross referenced with the spot numbers in the numerical output table and so will generally not be used by the user 23 outlined above it is possible and often preferable to automatically import data from a whole run of FindSpots directly into a Microsoft Excel spreadsheet This is often preferable a
25. can be helpful to first run the analysis and examine output with no or very low minimum value set Next examine some typical images If too Note that for this and the other automatic thresholding methods the same threshold is used for all Z sections within a stack but if the file also contains a time series a new threshold using the same parameters will be calculated for each point in the time series 12 many spots are detected with the current parameters one can then increase the minimum spot volume and run the analysis again It can be helpful to obtain the spot reference numbers see below of some of the true expected spots and examine the numerical output to determine their volume so as to have a better idea of the pixel volume of the expected spots The minimum spot volume should then be set somewhere below this value but not too close to it so as to leave enough room for variation in spot volumes 11 Setting the spot finding parameters in practice As can be seen from the preceding section a critical parameter to set for a FindSpots analysis is the threshold all contiguous pixels whose intensities are above the specified threshold are considered to be part of one spot Since the most useful threshold methods will be RelativeToMean and RelativeToGeometricMean this in effects means that the critical parameter to set is ThresholdValue or n above The greater n is the higher the threshold will be and the fe
26. es010905 test 1 2 and 3 1 images Link http bonedancer openmicroscopy org uk perI2 serve pl Page O0ME Web DBOb jDetail amp ID 12 amp Type OME Dataset Clicking on Dataset under the Search menu shows this view with thumbnail illustrations of a few images from each dataset Click on the name of a dataset to proceed to a detailed view of the chosen dataset You will be presented with a view like this Welcome David Schiffmann Open Microscopy Environment 4 No recant project creata naw Most recent dataset UsIRNA201005 Popup OME Projact Dataset UsiRNA201005 Dataset SS Id 20 Owner David Schiffmann Group OME Projects Search Locked False More info Projects Name UsiRNAZOT005 Datasets Description Save U20S cells un treated stained for Bub1 528nm BubRi 617 nm ACA CREST 685 nm DAPI Can visualise BOTH Bub1 and BubR1 in same cells Images Module Executions Chain Executions Other Annotation o 3 7 2 z 3 3 3 Your Current Annotation Save Mark Invalid View all 0 Annotations Annotate Images Search by Annotation Import Spreadsheet Import Export Image s Create a custom annotation of Select a Semantic Type Find Spots Import Modules Execute Chain View Chain Results To cluster thumbnails by Category select a CategoryGroup No CategoryGroups are used by this Dataset Can t find what you want in that l
27. et to use is Get Chain MEXes A summary of the steps to follow are provided by the Help button there First enter the dataset name in cell B1 When entering the dataset name into the Excel worksheet it is critical to enter the dataset name avoiding any trailing leading spaces these may inadvertently be picked up when one copies pastes the dataset name from the web browser Otherwise Excel will not be able to transfer the data from OME because the filename Excel is using will not exactly match the one in OME due to these extra spaces Next click Get Chain executions CHEXes A list will appear to the right of this button showing all the CHEXes Each one corresponds to a different run of Find Spots The more recent the FindSpots run the greater the ID number i e the lower down the spreadsheet document it is Click in column H on the ID number corresponding to the FindSpots run you would like to extract data from Then click the Get MEXes button Then as it says click on Get All results to get all the results or select a subset of them column E IDs and click Get results for selection This may sound complicated but should be clearer after trying it a few times Then wait whilst it queries the OME server and generates the spreadsheet This can take from a few minutes to several hours or days potentially for extremely large datasets depending on the number and nature of the images in the dataset and but one s pers
28. he spots found by FindSpots Make sure by examining a couple of output images with the FindSpots overlay that the parameters chosen including the number n of standard deviations above the mean yield output images that show relatively well separated spots here kinetochores i e that the actual spots in the image are not all merged into a few very large blotches which can happen if n is too low which will give very misleading data If on the other hand n is too high then only very small fractions of each spot will be found and some spots will not be found at all which can also give misleading results A complementary approach to visually examining the images is to browse the data tables within OME or download the data to a spreadsheet see later in order to count the number of spots found This enables one to check that this number is biologically sensible given the type of samples being studied However it is worth bearing in mind that there is no perfect thresholding Young et al 1998 and an acceptable result is somewhat subjective depending to a large extent on what is to be done with the data subsequently The simplest test of adequate thresholding is to see that the results are sensible in terms of where the found spots are and the number of spots found Note that in some instances what appear to be two separate spots may be identified by the algorithm as a single spot this can occur if at any point in the image stac
29. ible to fully delete images from an OME system using OME s command line administration tools which are beyond the scope of this document see http www openmicroscopy org system admin Or if one wants to run an analysis on just a sub set of the images that one has already imported into OME one can create a new dataset and use Add Images to add only the desired images to this new dataset these images can come from any number of different datasets This can be useful for example to save time by only analysing a few images whilst trying out different FindSpots parameters or because one is in fact only interested in a sub set of them perhaps because one has imported some unwanted files by mistake 9 Using the viewer A large part of the interaction with OME FindSpots is via the web based OME viewer Images displayed via the web interface of OME are generated using a plug in called Adobe SVG viewer which needs to be installed in order for images to be displayed If this is not already installed a message will appear explaining this with a link to download it The viewer has a number of floating panels to control the display The largest on the top left has sliding controls enabling one to move through the different Z sections and through time if the image consist of a time series either one at a time white arrows or automatically shaded arrows Note that before using the shaded arrows it is helpful to click on prel
30. ic threshold value for that image but in practice this is not critical Try to avoid images that contain large blotches of signal e g due to cell debris or precipitation of fluorophore containing material or where possible crop these areas out or avoid those Z sections Otherwise note which files this occurs in as it may be necessary to alter the method used for determining background in such images Re save them with the same filename as originally but in the corresponding sub directory in the after cropping directory 5 Connecting to OME This and subsequent steps will require use of a web browser For more details about how to access OME using a web browser see the OME installation instructions available from the OME website Note that if using a Mac it is recommended to use Apple s Safari web browser available as standard in Mac OS X for Windows Internet Explorer is recommended Connect to an OME server and log in 6 Importing images into OME Before any analysis can be performed on the images they must be imported into OME Create datasets A fundamental unit within OME is the dataset This is a group of one or more still images or movies Analyses such as FindSpots can be performed on whole datasets at a time so it is generally convenient to place all images to be directly compared against one another in a single dataset Datasets can only contain a single copy of any one image that has been impo
31. igned for the detection and analysis of objects within microscope images in particular fluorescence microscopy images The aim of this document is to provide guidance about installing and using FindSpots In the following sections we describe how to prepare images for analysis in FindSpots set up and run a FindSpots analysis import the data into Microsoft Excel for further analysis NB Installation of OME and access to the OME server Please note that this is explained elsewhere see the OME installation user guide at http www openmicroscopy org howto index html and the README file accompanying the Macintosh OME installer see http www openmicroscopy org install mac osx html Note As an example of the issues to be considered we describe here how to apply FindSpots for the analysis of kinetochore proteins in images of kinetochore labelled cells However the general points made are applicable to other types of image as well An online getting started tutorial and overview of OME can be found at http www openmicroscopy org getting started 2 Image formats Any image file imported into OME can be analyzed with FindSpots OME is able to import a large number of file formats Currently supported formats can be found at http www openmicroscopy org getting started import html FindSpots can be used not only for fluorescence microscopy images but also for electron microscopy images and any other microscope images th
32. import and the correct dataset has been selected click on Import files and file import will begin To check progress click on Tasks under Options at the bottom of the menu Note that it is not possible to import a file that has already been imported into the current OME installation even if the filename or location is changed This is because a digital signature is created for each file based on its contents This is a feature designed to maintain database integrity otherwise the same file may in practice be imported multiple times thereby losing the connection between all the actions performed on what is actually the same image It also saves storage space Instead to add a file that has already been imported one should instead use the Add images command to add the image to a new or previously created dataset described below Alternatively one can slightly modify the image e g slightly crop it which will alter its corresponding digital signature and allow a copy of it to be imported gt This was chosen as it has benefits from a design and database integrity point of view OME can also be run via the Unix command line and it is possible to use this to import a file more than once This more advanced method of interacting with OME is beyond the scope of this manual Please contact the developers for more information on this 7 Choosing a dataset If one has created more than one dataset it may be neces
33. ist You may want to Search or Create a new one Poe Eo Module Executions Chain Executions More info Define a protocol Tasks Logout Analysis chain Find and track spots was executed Analysis chain Find and track spots was executed Analysis chain Image server stats was executed against dataset UsiRNA201005 on 2005 08 24 15 19 16 by David against dataset UsiRNA201005 on 2005 08 24 15 19 16 by David against dataset UsiRNA201005 on 2005 08 24 15 19 16 by David Schiffmann The ID is 52 Schiffmann The ID is 51 Schiffmann The ID is 50 Summary and description of the selected dataset At this point one can start a FindSpots run by clicking on Find Spots under the Analysis menu or one can click on More info for more information about each image including results from previous FindSpots runs An alternative way to reach the image detail page is to click on the upper left corner of an image thumbnail displayed in any OME page Clicking anywhere else on the thumbnail will launch the image viewer for that image 8 Pre flight quality control Before running FindSpots it is recommended to check what files have been imported Ensure all files have been imported as intended NB After importing images but before running FindSpots it is recommended to compare what is in the actual OME dataset with the list of pre processed files that one intended to include in there Check if any were omitted from the imp
34. k there are pixels bridging the gap between the two spots Note too that in the overlay view the absence of a blue circle around a spot does not mean that it was not picked up but may simply mean that the centroid of that spot is not in the particular Z section being viewed a simple way to check this is to move up or down to neighbouring Z sections or switch the viewer to display all Z spots together If the dataset contains too many images to make it feasible to check them all then a representative selection should be tested 20 Inspection of numerical data To view the numerical results for a particular image go to the detailed view of data for that image e g by clicking in the top left of the image s thumbnail or by clicking Info when using the viewer to view that image As described earlier on the top right are displayed all the module executions performed on that image and the dates they were performed In this list are the FindSpots modules and the dates they were run Click on the run to navigate to a Module Execution Detail view There may be more than one run displayed if FindSpots has been run more than once on that image as mentioned above the correct one can always be verified by its data and parameters A window such as that shown below will appear Trae a j z Open Microscopy Environment v H iaceat Pront creat new recent dataset UsIRNA201005 Popup OME Home gt gt Module
35. ll appear in the worsheet if login was successful The SessionKey looks like a bunch of random letters and numbers Note that if you do not import any data for a while this step will eventually need to be repeated as one will effectively be logged out due to inactivity as a security measure M OME Excel xis gt A B Cc D E F G Login to OME 1 ES Set the host name of the OME server you use set your OME username and save this worksheet for ft 2 OME Server Name localhost After saving your connection settings click the Login to OME button and type your password 3 OME Username myOMEusername If your login was successful the OME SessionKey will contain a random looking string of letters and numbers 4 OME SessionKey 5 6 z 8 9 10 OM 11 12 Log into OME 14 15 Server URL http localhost perl2 serve pl Page OME Web Login 16 2 Username yOMEusername 19 20 Password 21 Cion 23 I Get Chain MEXes Get Image MEXes SampleQueries OMEconnection OMEquery J Importing FindSpots results into Microsoft Excel logging in to OME Using the OME Excel spreadsheet available from the OME website select the OME connection worksheet enter the server address and username and click Login to OME A window will appear in which one can enter one s password after which the session key will automatically be updated in spreadsheet cell B4 25 Importing data into Excel The next workshe
36. mportant notes before using the spreadsheet There are a number of important points to bear in mind before using the spreadsheet to import data from OME Before running the macros within the spreadsheet save it under a different name otherwise it will over write the existing contents of the file as it goes along which one may wish to keep Unless one is using the original version of the spreadsheet as downloaded from the OME website it is also important before running the macros to delete all worksheets corresponding to specific images that may be present from the last time it was run To do this select the first worksheet move to the last use the arrows in the worksheet bar at the bottom left of the window and shift click then go to Edit menu and choose Delete sheet s and click OK This is important because even if the resulting spreadsheet is saved with a name specific to your data the file may to begin with contain worksheets on it from older analyses which would be confusing at the time of subsequent analysis if they were not first deleted Ensure macros are enabled on your Excel installation to do this one may need to temporarily lower the Excel security level see the online Excel Help for information about this Excel will likely issue a warning when opening the OME Excel file saying The workbook you are opening contains macros Some macros may contain viruses that could be harmful to your computer Click
37. nd simpler especially when one needs to obtain and process data from many images in addition the tables thus generated contain a more complete set of data than can be obtained by using the merge tables option Importing a run of data into Microsoft Excel allows complex further analysis to be performed In this context a run means a set of images in any one dataset that has been analysed by FindSpots any number of images can be analysed together in this way within the limits of the OME server and Microsoft Excel In order to import data into Excel it is necessary to first open a copy of a special Excel spreadsheet that contains the macros for importing data from FindSpots This is distributed as part of the Macintosh installer for OME available via http openmicroscopy org install mac osx html look for the Excel spreadsheet file called OME Excel xls in the EXTRAS folder By running the macro as described here a spreadsheet file will be generated containing the FindSpots data for the specified dataset and the particular run of FindSpots on that dataset chosen by the user The spreadsheet file will contain several worksheets with one worksheet per image file each containing information and output data for the FindSpots output run on that image file This includes the image filename the emission wavelengths and a table of statistics for each spot in that image including spot total intensity mean and position in 3 D space I
38. oad panes this pre loads the data for the various Z sections or timepoints yielding a smoother transition between them There are a number of other panels as well Info opens up a page in one s web browser with links to further information about this image e g the different FindSpots analyses performed on it Stats displays basic statistics useful for example when establishing parameters for automated thresholding see below The Color map floating panel allows one to change the black level and white level of the image to switch wavelength channels on and off in the image and to choose which emission wavelengths to display Note that it is possible to zoom in and out when viewing an image If you re using the Adobe SVG viewer for the Mac with the Safari web browser the recommended browser to use when using a Macintosh then gt For more information on this tool see http www adobe com svg main html 10 apple click will do a zoom in apple shift click will do a zoom out and option drag will allow you to move the image window around The Adobe SVG plug in for Windows works slightly differently see http discover nci nih gov mim S VGHelp html The SVG viewer is limited in that it only allows you to look at 3 wavelengths at a time because each wavelength is assigned to one of the RGB red green blue channels However you can click on the wavelength s label usually the nominal emission wavelength sometimes the fluo
39. oggle between a display of spots only in the current time point or a display of spots at all times 18 v Z slice 31 55 lD lt p Timepoint 1 1 lt I gt lt p Channel ron White e Example of FindSpots output overlaid over the original image One optical section of the image is shown at a time The blue circles show spots whose centroids are in the particular optical section currently displayed Note the above image contains data from four emission channels however the SVG viewer used here can only display 3 at a time to change one of the channels to display a different wavelength click on the wavelength and select the new one from the pop up menu see the red boxed area above One can toggle between a colour and a greyscale representation by clicking on the Color Map or Greyscale Map headings shown next to the image One can also turn channels on and off as desired It can sometimes be easier to change the display mode from colour to grayscale in order to more clearly see the blue circles outlining each spot that has been found as shown below 19 HOO 341ABR019_R3D_D3D dv OOD 341ABR019_R3D_D3D dv EE A Greyscale Map TA Z slice 24 46 Ti t 1 1 Channel Black White Black White raed lt gt al oe 14 lt gt a Level Levei a iS Level Image Info Stats Save settings SaveasTIFF Preload planes Overlays Ke 2 Changing from a colour to a greyscale display can help to highlight t
40. onal computer can still be used whilst this is happening 26 27 09 05 CenpBsiRNAB270905 x ls gt A B Cc D E F G H I J K l Dataset 27 09 05 CenpBsiRNAB270907 Hel 2 Chain Find and track spots elp 3 Edit Dataset and Chain name then A Get Chain E r CHEX CHEXes 5 et Chain Executions es Experimenter id Timestamp Id 6 1 20 9 05 14 54 20 7 Module Find spots 1 20 9 05 14 54 21 8 Edit Module name Select a CHEX ID from the 9 Id column in the CHEX table on the right then 10 11 Get MEXes 12 13 Then Either 14 15 Get All Results 16 Or select one or more MEX Ids and Get Results for Selection 20 Construct a summary formula for an image using one or more 21 worksheet references to one of the result worksheets for FindSpots 22 Place the formula at the top on an empty column in the MEX table below 23 Select the formula and click Fill down formulas 24 to generate summary formulas for each image Fill down Formulas 26 MEXes For CHEX ee R E an i4 4 gt gt I Get Chain MEXes Get Image MEXes SampleQueries _ OMEconnection _ OMEquery Importing FindSpots results into Microsoft Excel How to specify the desired dataset and FindSpots run to import data from 1 Go to the Get Chain Mexes worksheet enter the dataset name and click Get Chain Executions CHEXes The worksheet will look similar to that shown here In columns F H will appear a list of ID numbers each c
41. orresponding to one FindSpots run Select one of the cells in column H corresponding to the desired run and click Get MEXes 21 eae 27 09 05 CenpBsiRNAB270905 xIs gt A B G D E F G EI I J K S l Dataset 27 09 05 CenpBsiRNAB270907 Hel 0 2 Chain Find and track spots p 3 Edit Dataset and Chain name then Get Chain Executions CHEXes CHEKes 5 Experimenter id Timestamp Id 6 1 20 9 05 14 54 20 7 Module Find spots 1 20 9 05 14 54 21 8 Edit Module name Select a CHEX ID from the 9 Id column in the CHEX table on the right then 10 11 Get MEXes 12 13 Then Either Get All Results 16 Or select one or more MEX Ids and Get Results for Selection 20 Construct a summary formula for an image using one or more 21 worksheet references to one of the result worksheets for FindSpots 22 Place the formula at the top on an empty column in the MEX table below 23 Select the formula and click Fill down formulas 24 to generate summary formulas for each image a Fill cown Formulas 26 MEXes For CHEX 20 es 27 Experimenter id Image id Status Timestamp Id 28 1 105 FINISHED 29 9 05 17 39 1416 29 1 104 FINISHED 29 9 05 17 39 1419 30 1 103 FINISHED 29 9 05 17 40 1422 31 1 102 FINISHED 29 9 05 17 41 1423 32 1 101 FINISHED 29 9 05 17 41 1425 33 1 100 FINISHED 29 9 05 17 42 1427 34 1 99 FINISHED 29 9 05 17 43 1429 35 1 98 FINISHED 29 9 05 17 44 1432 36 1 97 FINISHED 29 9 05 17 45 1434 37 1
42. ort This is particularly important if one intends to import the resulting data into Microsoft Excel Because of the way the final output is generated in Excel by the import macro it is much simpler to run the FindSpots analysis and the subsequent import into Excel with all the relevant images files in place on the OME dataset rather than add and process them in stages later on Using the Add images button Images that have already been imported into OME can be added to a dataset using the Add images command shown in a later screen shot The Add Images and Remove Images buttons are located next to the dataset name and is an option only available in the dataset detail page which one obtains by clicking on a dataset name from any other page such as a search page Add images has several potential uses If one would like to add certain images to a dataset which have already imported into another OME dataset one can use the Add images button to add these to the dataset being viewed without having to re import them Indeed as mentioned above it will not actually be possible to re import images if they have already been imported into OME Any images added in this way continue to also be in the other dataset they were in unless they are removed from it Similarly images removed from datasets using the Remove Images command continue to exist in other datasets or unattached to a dataset Currently it is only poss
43. ram Before downloading the data it is possible to create larger merged tables by selecting some of these outputs at the top of the page and clicking the Merge Outputs button this will put those outputs into the same table for easier viewing This table has a row for each spot in each channel in your image Click the Download as text link to get a tab separated text file with that data The first time this is done it will probably ask if you want to save the file or open it with a certain program If you tell your browser to open these types of files text tab separated values MIME type in Excel clicking the Download as text will just open them up in Excel from then on 14 Identifying spots by number Each spot found by FindSpots sometimes referred to in FindSpots as a feature has a unique reference number assigned to it which can be used to look up the detailed statistics position total intensity etc about that spot in the output tables In order to view this reference number click on one of the blue circles and a small floating window will appear in the viewer as shown in the screen shot below The important number here is the Name in the numerical output table a Feature number is listed for each 22 spot see later which is identical to the Name shown here and enables one to look up statistics for the spot Once this window has been opened one can simply move the pointer to other spots i e to other
44. rchantability and fitness for a particular purpose are disclaimed In no event shall the National Institutes of Health University of Dundee or their affiliates be liable for any direct indirect incidental special exemplary or consequential damages including but not limited to procurement of substitute goods or services loss of use data or profits or business interruption however caused and on any theory of liability whether in contract strict liability or tort including negligence or otherwise arising in any way out of the use of this software even if advised of the possibility of such damage Although efforts are taken to protect the content of Open Microscopy Environment software and website against known viruses the National Institutes of Health and the University of Dundee cannot guarantee that the material and downloads on the website are free from all malicious code It is your responsibility to ensure that you have taken precautions to protect your own computer system from virus infection To the extent permitted by law the National Institutes of Health and the University of Dundee will not accept any liability arising from any errors in or unavailability of the Website or documents located therein The names of actual companies and products mentioned herein may be the trademarks of their respective owners 1 Introduction FindSpots is a module within the Open Microscopy Environment http www openmicroscopy org des
45. resholding methods which are the three methods most likely to be useful for analysing a series of fluorescence images The choice between RelativeToMean and RelativeToGeometricMean will depend on the images in question although RelativeToGeometric Mean is expected to give a more robust measure of threshold in a recent study of kinetochore labelled cells we found that RelativeToGeometricMean did not give satisfactory results and used RelativeToMean instead The concept behind RelativeToMean and RelativeToGeometricMean is that they provide a simple but effective objective way to determine a threshold that can at least to some extent take account of varying background levels for each image or more correctly each stack of optical sections Although Absolute allows one to set an absolute threshold it is rarely of use because images will typically vary in background level and it is desirable to vary the threshold on an image by image basis to reflect this rather than using a fixed threshold level Minimum spot volume The other parameter which needs to be set is the minimum spot volume measured in pixels This is important as by setting it sufficiently high it can in effect filter out unwanted spots e g those due to random noise or non specific antibody labelling from the output It should of course not be set too high otherwise spots of interest may also be lost In order to choose a value at which to set this parameter it
46. rophore in the color map and you will get a popup menu of all wavelengths collected so that you can assign a different wavelength to one of the RGB channels This can be seen in the screen shot later showing a graphical overlay of FindSpots results The Java based viewer in Shoola a Java based interface to OME which does not yet support FindSpots is able to blend the four wavelengths in more ways so that you can see them all at once 10 Setting the spot finding parameters theory In order to use FindSpots to detect and measure objects within an image it is necessary to set a number of parameters These are all set using the form shown below Open Microscopy Environment v 4 fn recur project create nev nt pr e ataset UsIRNA201005 Popup Home nalysis gt Find Spots Project Selected dataset UsiRNA201005 Enter parameters for findSpots Projects Datasets Time From to Begining _ Timepoint Images m Module Executions end OTimepoin Chain Executions Channel 685 ThresholdType RelativeToGeometricMean_ Create a Template ThresholdValue 3 Annotate Images Search by Annotation Min Spot volume 30 Import Spreadsheet Run FindSpots Import Export Image s F indSpots parameters window Set the parameters for the analysis and click Run FindSpots FindSpots defines a spot as any contiguous set of pixels above the threshold value that are larger than a minimum size pre set b
47. rted but images can be listed in more than one dataset see Add Images later Datasets can be created in advance of image import by clicking on Dataset under the Create menu the list of menus is displayed on the left hand side of the screen However you may find it simpler to create datasets as needed during import To import images click on Images under the Import menu One can choose whether to import files into an existing dataset or a new dataset To import files into an existing dataset click the Existing dataset button then select the dataset name from the pop up menu adjacent to it Alternatively to create a new dataset click the New dataset with name button and enter a name for the z Whereas it is often possible if the spots are small enough compared with the image as a whole to estimate background using the mean of the whole image this may not be valid if there are large blotches or inhomogeneities of colour in the image in which case the mean should be estimated from manually selected normal regions of the image dataset in the box to the right together with a more detailed description if desired in the box below When choosing the dataset name avoid commas ampersands amp or symbol marks e g etc as this can cause problems when importing the FindSpots data subsequently into Microsoft Excel If these have already been used in the dataset name try to rename the dataset within OME
48. sary to switch between these Indeed before running FindSpots it is necessary to first open the dataset on which FindSpots is to be run Welcome David Schiffmann Open Microscopy Environment v 4 No recent project create new t recent dataset 06 07 04 05 140405 335 amp 341 Bub1 ACA Popup a Hone Project 06 07 04 05 140405 335 amp 341 Bub1 ACA Preview 58 Project Preview 1 p xi No Project ne a La 2 DO Projects ca Images Module Executions Chain Executions Other La Annotation Create a Template a z annotateimages Welcome to the Open Microscopy Environment Search by Annotation Import Spreadsheet ST Most of your initial tasks with OME will start with this page the Home page From here you can create new projects and datasets as well as import images For more sophisticated tasks you can nativigate to various New Project New Dataset Import Images Recently Imported Images Recently Executed Chains Import pages using the menu on the left or using the links given to you in the previews above If for some reason you get lost you can always return to this Home page by clicking the OME logo in the top left hand corner of Export Image s your screen Find Spots Import Modules Execute Chain View Chain Results Define a protocol Tasks Logout OME home page after logging in At any time at the homepage or elsewhere clicking on Dataset under the Search
49. sections may need to be removed e g to remove additional underlying cells It is likely that this step can be accomplished using the software supplied with the microscope system for example in the case of Applied Precision s DeltaVision microscopy system one can use their softWoRx software running on Linux this is a software package principally for image deconvolution of DeltaVision files and for the analysis of DeltaVision files and other image formats Alternatively one can use stand alone image processing analysis packages as appropriate for example softWoRx Explorer a Java application also by Applied Precision Inc for Mac or Windows http www api com lifescience softworxexplorer html or ImageJ a freely available package for image processing and analysis http rsb info nih gov ij for which many plugins with advanced features are available When dealing with deconvolved files a simple but helpful step is to prepare first a before cropping directory containing sub directories with just the original deconvolved images for any one experiment i e before any cropping is performed this will make it easier to select the right ones to crop instead of manually scrolling through large directories which contain also the non deconvolved images too as well as log files and minimises the risk of opening and working on a non deconvolved image by mistake It is also recommended to prepare a separate after cropping directory
50. tly click Run FindSpots Immediately after this one can repeat this to queue further analyses tasks on the same dataset or choose other datasets and queue FindSpots analyses on them Or one can perform other tasks in OME such as importing further images Note Do not attempt to add an image to a dataset by adding one from an existing dataset or by importing an image s whilst FindSpots is running it will cause unpredictable and potentially serious problems to the software Checking progress At any time it is possible to check on the progress of an OME task such as image import or a FindSpots run by choosing Tasks from the Options menu The status will display FINISHED when the task has completed or an error message will be shown in the task was not completed 15 Tasks in the current user session e ff Refresh Home 2 AutoFill Print ail Apple Support Apple Store mac macosx Microsoft MacTopa Office for Macintosh MSN Blo a samones z Other S 2 Projects ry Datasets Images S Module Executions Chain Executions w Other g i Crea Annotate Images Search by Annotation Import Spreadsheet A x 2 Export Image s Find Spots Import Modules Execute Chain View Chain Results Define a protocol Logout Open Microscopy Environment z4 Home gt Options gt Tasks
51. ton This will display means geometric means sigmas standard deviations and geometric sigmas for all of the channels in the image To get a more stable threshold value take the absolute threshold determined earlier through moving the sliders around as described above and convert that number to some number n of sigmas above the mean or geosigmas above the geomean i e n absolute threshold mean sigma Then use this parameter it does not have to be a whole number as the parameter referred to as ThresholdValue for a statistical threshold i e RelativeToMean or RelativeToGeometricMean Worked example Let us assume that for a particular image by adjusting the sliders a threshold of 1205 seems optimal for the chosen wavelength channel in the image In the Statistics window look up the mean and standard deviation called sigma of the pixel intensity for the image Let us assume that the mean is 503 and the standard deviation is 223 Calculate how much the threshold is above the mean 1205 503 702 Calculate how many standard deviations this value 702 corresponds to 702 223 3 5 approx Hence for this image the Threshold Value setting should be 3 5 It is recommended to check a few other typical in the dataset to ensure that a Threshold Value of 3 5 will give similar results To do this open another image and view the Statistics window for that image let us assume that the mean is 398 and standard deviation is 18
52. wer and or smaller the spots will be Visualising the effect of threshold In order to know what to set ThresholdValue to it can be helpful to visualise the effect of different threshold values on the image prior to running FindSpots To do this first open the image in a suitable viewer although this need not be a viewer within OME it may be convenient to use one of the OME viewers for this To open an image using the standard viewer described above click on one of the images in dataset to open the viewer clicking anywhere on the small thumbnail view of an image will open the viewer except for the top left corner which will open a page with information about the image and commands that have been run on it The aim here is to set the white level and black level to the same value so that you get a binary image i e pixels are fully on or fully off Displaying the image in greyscale and selecting the proper channel for the greyscale display is one way but you can do the same thing with the red green or blue channels To set the white level and black level to the same value set the black level first with the slider so that the spots you want to find are visible but the background is not Now click above and to the left of the black level slider position i e in the white level slider area to the left of the black level you set The white slider will jump and set itself to the value of the black slider What you should see in the display
53. y the user prior to the run An important parameter to set is therefore the threshold above which a pixel can be considered part of a spot 11 This can be fixed by the user the Absolute threshold option within FindSpots or it can be calculated automatically by FindSpots on an image by image basis for each image within the dataset being analysed FindSpots includes a number of algorithms for calculating such an automatic threshold and it is critical to choose a method and associated parameter s if relevant that is suitable for the images being analysed Note that in any one run of FindSpots all images will be analysed by the same thresholding algorithm These automatic algorithms are RelativeToMean where threshold mean of all pixels in stack n standard deviation of all pixels in stack RelativeToGeometricMean where threshold geometric mean of all pixels in stack n standard deviation of all pixels in stack FindSpots also has implementations of four other algorithms for automatically determining a threshold Otsu Kittler MomentPreservation and Maximum Entropy but these are generally not of use for typical fluorescence microscope images and will give far too low a threshold value to be of use for such images The parameter n above is referred to within FindSpots as ThresholdValue and is only relevant for RelativeToMean RelativeToGeometricMean and Absolute methods altering it will have no effect on the other th
Download Pdf Manuals
Related Search