biofilter - Biological Knowledge Integration Utility
Contents
1. VARIATION FILENAME variations bn SETTINGS_DB filename SETTINGS DB bio settings cn MAX_GENE_COUNT integer MAX GENE COUNT 30 SNPS_SOURCE filename SNPS SOURCE projects ritchie biofilter Affy6 0 v27 txt INCLUDE_GROUPS group group group INCLUDE GROUP FILE filename INCLUDE GROUP FILE group list txt MODEL FILENAMIE filename MDOEL FILENAME affy50 models bin DISEASE DEPENDENT string DISEEASE DEPENDENT alz bio knowledge txt POPULATION string POPULATION NO LD MODEL BUFFER INIT integer MODEL BUFFER INIT 10000 MODEL BUFFER MAX integer MODEL BUFFER MAX 100000 MAX GENE COUNT integer MAX GENE COUNT 30 Ritchie Lab 7 7 7 biofilter Reference PREFERRED_ALIAS filename 9 PREFERRED ALIAS my favoriate genes txt 9 LOAD_ALL_ALIASES YES NO 9 LOAD ALL ALIASES NO 9 REPORT PREFIX string 9 REPORT PREFIX data bio alz 9 IMPLICATION IDX DUPLICATE WEIGHT float 9 IMPLICATION IDX DUPLICATE WEIGHT 0 25 9 HTML_REPORTS YES NO 9 HTML REPORTS YES 9 DISEASE DEPENDENT LEVEL ALL MODELS GROUP LEVEL DD ONLY 9 DISEASE DEPENDENT LEVEL DD ONLY 9 COLLAPSE ASSOCIATION REPORT YES NO 10 COLLAPSE ASSOCIATION REPORT YES 10 BINARY MODEL ARCHIVE YES NO 10 BINARY MODEL ARCHIVE YES 10 ASSOCIATION REPORT YES NO 10 ASSOC2. Ritchie Lab Software biofilter Reference 26 Generating Sample Configuration Users can use biofilter to generate a skeleton sample configuration Most command line options will be incorporated Users should edit this configuration with appropriate settings Filenames follow standard unix convention for loca tion This allows them to be expressed with full paths as well as paths relative to the directory in which the applica tion is run this includes just using the filename if the file exists in the same directory as the application was run Fields that have no default value will be commented out in the example configuration In some cases such as SNPS_SOURCE a value is required before certain types of execution can be made biofilter sample config gt sample config Variations data VARIATION_FILENAME variations bn BioFilter data SETTINGS_DB bio settings cn Max number of genes before we ignore the group MAX_GENE_COUNT 30 The source file for the RS numbers in your dataset SNPS_SOURCE List the various groups by group name separated by spaces INCLUDE_GROUPS Set the filename for the output model list none writes to std out MODEL_FILENAME NONE Set the initial size of the model buffer MODEL_BUFFER_INIT 10000 Set the upper limit to the buffer Bigger gt faster but must remain within the limits of the hardware or could cause the application to fail or become so slow that it will
3. 4 Introduction to makin 47 PythonSpeed Perform 47 http drp rubyforge RSID Chrom Region Name 10000010 4 ENSG00000185774 80333 KCNIP4 10000023 4 ENSG00000138696 658 BMPR1B 10000030 4 ENSG00000138821 64116 SLC39A8 10000092 4 ENSG00000185774 80333 KCNIP4 10000160 4 ENSG00000198092 389208 FLJ16046 TMPRSS11F 4 ENSG00000215127 401135 SYT14L 10000169 4 ENSG00000138771 57619 SHROOM3 10000185 4 ENSG00000184985 57537 SORCS2 1000022 13 ENSG00000125246 171425 CLYBL ENSG00000163629 5783 PTPN13 ENSG00000145283 345274 SLC10A6 10000255 4 ENSG00000168843 56884 FSTLS 10000226 Ritchie Lab Software biofilter Reference SNP SNP Model Report The SNP SNP Report lists the details for each model from a previously generated model file Details include e Left Right SNPs that make up the model rs2072539 rs1990310 rs2072539 rs1805488 e The Group s in which the pair of Loci were found rs176590882 rs17701871 e The Genes from which each of the SNPs were found When possible each of elements reported will reflect the highest form of the name provided by the user At the very least Ensembl IDs will be used However if the user specifies gene Aliases those will be used preferentially over the less familiar Ensembl ID The excerpt to the left shows an example of SNPI Gene Groups SNP2 Gene Groups 2072539 GRIN2B KEGG 05010 1805488 GRIN2B KEGG 05010 2072539 GRIN2B KEGG 05010 1990310 ENSG0000013
4. Aliases This file provides one or more common gene names which can be substituted for Ensembl IDs in reports Aliases must be alphanumeric no spaces and must match an alias known to the biofilter Known aliases are those that were found in Ensembl s external synonyms associated with EntrezGene amp Uniprot TrEMBL and Swiss Prot Currently only aliases which map back to a single Ensembl Gene ID are used This file is present only as a convenience for the user and is optional Example file NMT1 FURIN NRD1 S100B ATP2A2 SNPS_SOURCE The SNP Source file contains all SNPs to be used in the analysis Generally this will match the SNPs from the plat form to be used in the analysis However it is also possible to use a highly restricted set for other types of analysis such as identifying which genes a set of interesting SNPs might be found in The format is very simple List all RS IDs in their integer format Each ID should be separated by whitespace 10000169 10000185 10000201 1000022 10000226 1000025 10000255 10000266 Disease Dependent Genes Ritchie Lab Software biofilter Reference 11 Users can tag genes as being disease dependent and add grouping information Assigning genes to disease depend ent groups causes two things to happen First it allows the user to increment the implication index of disease dependent genes for every disease dependent meta group they assign it to Second by pr
5. Ensembl ID for all gene names This ID is stable and has a single meaning However most users will prefer to see IDs listed in more familiar terms Providing a list of familiar aliases allows the application to use a known synonym for the gene without it having to make any assumptions for the user For more information on the format of this file please see the documentation on gene aliases Lgene aliases txt For the purposes of example we ll use a really small one with some genes we know will be in our report This file is called gene_aliases txt To use these aliases edit your configuration file and change the line containing PRE FERRED_ALIAS Remove the from the beginning of the line and add the filename to the end of the line It should look something like what you see below User can specify aliases for genes the alias must be present in the database PREFERRED ALIAS gene_aliases txt Selective Search The biofilter contains a large amount of information much of which might be completely outside of the user s inter est Users can select as much or as little as they want to use We ll be restricting the search to include only groups that have Alzheimer s listed in the comments This is probably not a very good example since it will only capture a tiny amount of knowledge but for the purposes of our example it will do just fine Biofilter gives the user the ability to query it s built in group informatio
6. Finally we get the list of reports that were generated According to the list above we should find an Alias report this describes which genes were given an alias a gene gene model summary the actual gene gene model file and the snp snp model file tutorial model summary txt This file lists the contents of our disease dependent groups In this case there is only one group Since we only recog nize a subset of gene names we produce this list to help the user recognize that A We got the important genes B They correspond to the same gene that the user expected Adding new group 176647 alz assoc Genes Ensembl Start Stop Snp Alias ID ID Chrom Pos Pos Count AGT ENSGQ0000135744 98998 APH1A ENSGQ0000117362 96653 APOA1BP ENSGQ0000163382 97981 APOA2 ENSGQ0000158874 98229 CAMK1G ENSGQ0000008118 98856 CFH ENSGQ0000000971 98642 CHRNB2 ENSGQ0000160716 97575 CLCNKB ENSGQ0000184908 78893 228904897 228916564 8 148502512 148508156 Q 154828178 154830715 Q 159458707 159460042 1 207823668 207853906 12 194887764 194983255 16 152806881 152818975 1 16242834 16256390 7 The ID column 3 is the internal ID that we use with the configuration parameter INCLUDE_GROUPS Ritchie Lab Software biofilter Reference 35 tutorial gene gene 146491 78609 93756411419112 80048 8012641121419 88966 9431541121419 79109 89515411214112 79109 83530411214112 8683198258 41121419 80091 82408 411214112 78373 80377411219112 8347
7. can be found on a platform of interest By default the biofilter comes with 4 different platform files but the user can use any list they wish as long as the file contains only integer representation of the RS ids found on the platform Any file that can be used as a SNPS SOURCE file can be used as a coverage file The baseline coverage is based on the SNPs found using the regular SNPS SOURCE value or s This is treated as the total Additional coverage sources are displayed as additional columns Each entry shows the number of SNPs covered by that platform for a given gene A coverage report might look as follows biofilter sample config s Illumina 660Quad txt report gene coverage alz genes txt C Illumina 660Quad txt Gene Ensembl id Total T1Lumina 660Quad txt 593544 AGT ENSG 0000135744 78 78 APH1A ENSG 0000117362 11 11 APOA1BP ENSG 0000163382 5 5 APOA2 ENSG Q000158874 12 12 CAMK1G ENSGQQ000008118 22 22 CFH ENSG 0000000971 82 82 CHRNB2 ENSGQQ000160716 1 1 CLCNKB ENSGQQ000184908 18 18 The gene alias is shown in the first column followed by the ensembl_id The total represents the number of SNPs found in the SNPS SOURCE file and is followed by the SNP count for each of the coverage files Using the option detailed coverage lists all RS IDs associated with the genes and their position for the text report Ritchie Lab Software biofilter Reference 15 Additional coverages can be added using additiona
8. children found Total Gene Gene Model Count 171802 Gene Gene Model Summary Snp Snp Model Estimates Impl Idx Count i 67410870 2 4812558 3 69458 4 653 Snp Snp Model Generation Summary 930386 69363 653 Ritchie Lab Software biofilter Reference 34 The next portion describes the estimated number of SNP SNP models for each Implication Index This number is an overestimation since it doesn t take into account actual overlap between gene gene models To describe the output of our snp snp models we get a summary describing how many models were produced This number is the exact number that was produced If you do the math you will see that there aren t exactly 1 000 000 models produced This has to do with the fact that gene gene models produce varying numbers of snp snp models including some models that might be produced by another gene gene model biofilter keeps the requested model count in mind and tries to get close to the requested models These models will have 0 overlapping members and will represent the highest Implication Index possible In our case we only represented 930 386 out of the 4 812 558 possible models with an implication index of 2 0 The ones that were generated were simply the ones produced by the first N gene gene models Alias Report tutorial aliases html Gene Gene Model Summary tutorial model summary txt Gene Gene Models tutorial gene gene Snp Models tutorial snpsnp
9. or it can be used to produce a selective list of snp snp models Both of these files can be written as text or in a binary format The gene gene models are sorted by Implication Index such that those with the highest score are at the beginning of the file To produce a gene gene model file simply pass the argument W on the command line This option can take 2 op tional parameters Minimum Implication Index and the Maximum SNP SNP model count These determine the num ber and quality of SNP SNP models to be generated immediately upon completion of generating the Gene Gene model file The user must provide both values or neither For our purposes let s produce up to One million models with the worst Implication Index of 2 0 biofilter sample config W 2 0 1000000 Illumina 66 Quad txt 592652 SNPS eee eee eee eee 593548 matches in our database Group Group ID Group Count Gene Count Gene Ontonology 1 6305 10673 KEGG 2 204 4640 NetPath 3 21 161 pfam 4 3904 16261 Reactome 9 4024 675 DIP 2 1310 1045 Disease Dependent alz txt ALZHEIMERS alz assoc 176647 1 8 The first few lines just describe the state of our database with regard to the input we have provided Out of our 593 548 SNPs 592 652 were found in our local database We also are provided a list of meta groups and their various counts and IDs These Group IDs are useful when selectively using one or more meta groups The group and gene counts represent the various
10. rs2042370 rs1862710 rs4645989 rs1 52571 NDUFV2 rs4148964 rs11081459 rs4148965 rs1039825 rs1472944 rs977581 rs874250 rs4148966 rs4148967 rs4148968 rs12966444 truncated Associations NetPath Associations pfam Associations Reactome Associations DIP There is a lot there even though we included only 7 groups Many of the GO groups had one or more child groups and the KEGG group we added had a large number of genes associated with it 127 to be exact If we were to use all 127 genes to generate models the resulting model count could be tremendous depending on the numbers of SNPs in each of those genes If you look through the report as generated by the biofilter you will see that some have quite a few SNPs This problem brings up the value in one of the configuration options In the configuration file you will find a line similar to the following Max number of genes before we ignore the group MAX GENE COUNT3 Ritchie Lab Software biofilter Reference 31 While the comment might suggest that we ignore the group it s a bit misleading The application won t actually cre ate models with any group larger than 30 However it will attempt to traverse any child groups and consider produc ing models with those if they have 30 genes or less In the case of the KEGG group above there are no child groups so that group will not yield any models with our current setting Should you want to ensure th
11. which they prefer These aliases are used in place of the Ensembl ID in subse quent reports See formatting details for help creating the file LOAD_ALL_ALIASES YES NO LOAD ALL ALIASES NO Loads all region aliases gene names and generates a report report prefix aliases This makes it easier for the user to lookup the Ensembl IDs that are used by default in the reports REPORT_PREFIX string REPORT PREFIX data bio alz When a report is produced that is sent to file instead of std out it will use the value of REPORT_PREFIX as the first part of the file name Acceptable options can be anything that is acceptable for the filesystem except for whitespace i e don t use spaces or tabs IMPLICATION_IDX_DUPLICATE_WEIGHT float IMPLICATION IDX DUPLICATE WEIGHT 0 25 Disease dependent DD and disease independent DI groupings contribute differently toward the implication index For DD groupings only those groupings which produce the gene gene model are counted Each one counts one point For DI groupings we add a single point for each unique group where at least one gene is present For each DI grouping where both genes are present the IMPLICATION IDX DUPLICATE WEIGHT is added to the final score HTML REPORTS YES NO HTML REPORTS YES When set to yes most reports will be written in html format DISEASE DEPENDENT LEVEL ALL MODELS GROUP LEVEL DD ONLY DISEASE DEPENDENT LEVEL DD ONLY Users can choose to filter models based on their a
12. 1419 75837 91412 4 1 21419 75837 75847 4 M2s 64546 91412 4 1 21419 64546 75847 4 1 2 4 9 64546 75837 4 1 21419 Ritchie Lab Software biofilter Reference 22 Genes Report This is actually part of the gene gene model output and is a text file which contains the information required to asso ciate genes with their contents SNPs and their potential contributions toward implication index Each column is separated by a tab For compound fields such as SNP lists and group IDs constituent members are separated by a character There are 5 Columns not all will be filled Alias Gene ID SNPs Disease Independent groups Disease Dependent Groups TYMS 44549 596909 3786362 11540152 11540153 2 4 9 360684 360685 ENOSF1 44559 596909 3786362 11540152 11540153 1 4 360685 MEST 44568 596909 3786362 11540152 11540153 1 2 14 9 TWSG1 44857 3322110898 12680 11559053 28552921135867116 4 RALBP1 44861 3322110898 12680 11559053 28552921 35867116 1 2 4 9 PPP4R1 44868 3322 10898 12680 11559053 28552921 35867116 1 4 CHSTS 45509 417808 418546 1155514117694469 28693844 1 4 OXA1L 45871 1061040 1805059 1805061 2281677 8016634 8018462 1 2 4 SLC7A7 45894 1061040 1805059 1805061 2281677 8016634 8018462 1 4 9 MRPL52 45927 1805059 1805061 2281677 8016634 8018462 1 MMP14 45945 1061040 1805059 1805061 2281677 8016634 8018462 1 2 4 Ritchie Lab Software biofilter Reference 23 SNP Cleanup Report After loading the SNPs f
13. 4 ABCG1 rs4148083 rs4148084 rs4148085 rs9975740 rs4148087 rs1117640 rs4148088 rs4148089 rs41480 PSEN1 rs214273 rs8006497 rs36235 rs214260 rs165933 rs362377 ENSG 0000162736 rs19494342 rs16831846 rs12239946 rs6664438 rs6677637 ENSGQ0000143801 rs2073489 rs12956490 ENSG 09000167755 rs1654537 truncated Associations KEGG KEGG 5010 127 ENSGQ0000015475 rs18139 rs181396 rs181402 rs181405 rs9604787 rs181408 rs181417 rs5746474 rs5747351 rs9605401 rs738095 ATP2A2 rs3026445 rs3026457 rs1860561 PSEN1 rs214273 rs8006497 rs36235 rs214260 rs165933 rs362377 NCSTN rs10494342 rs16831846 rs12239946 rs6664438 rs6677637 PSEN2 rs2073489 rs1295649 BACE1 rs7083 rs522843 rs687740 rs473210 rs551662 rs676134 ENSGQ0000132906 C rs6685648 rs2020902 rs4646018 rs2042370 rs186271 rs4645989 rs1052571 NDUFV2 rs4148964 rs11081459 rs4148965 rs1039825 rs1472944 rs977581 rs874250 rs4148966 rs4148967 rs4148968 rs12966444 truncated Associations NetPath Associations pfam Associations Reactome Each relationship is nested with tabs on separate lines In the example above GO 0042987 has some child relationship to GO 0042982 The numbers beside a group ID indicate the number of genes associated with the group An optional setting COLLAPSE_ASSOCIATION_REPORT can be set to true to collapse groups where models would be generated This allows the user to see more clearly which genes will be combined to produce mod
14. 6 83629411419112 88362 93366411219112 79632 93449 321419 85454 87349 3419112 86798 97921311419 truncated This file lists all gene gene pairings and the number of models that the pairing yielded This is the actual model file and might be unreadable binary depending on the value of the configuration property BINARY_MODEL_AR CHIVE is set to YES The first line indicates the number of gene gene models contained int the file Subsequent lines contain the gene IDs their Implication Index and the Disease Independent information associated with the pairing The file is sorted by Implication Index then gene 1 then gene 2 No duplicate gene pairings should be found A more understandable report is the tutorial model summary txt file Gene SNP Gene SNP Impl ModelsGroups Name Count Name Count Index CountDI DD PAFAH1B1 8 GLI2 31 l 2481 PAFAH1B1 8 CHRNB2 l 2 81 176647 GLI2 31 CHRNB2 1 2 311 176647 GLI2 31 DRD2 18 1 5581 NR B1 a DRD2 18 1 181 NR B1 al GLI2 31 1 Slik TGFB3 6 TGFB2 22 1 1321 ROCK1 19 EZR 10 1 1901 ROCK1 19 ICAM1 5 1 501 ROCK1 10 MSN 2 1 201 ROCK1 10 VCAM1 22 1 2201 ICAM1 5 EZR 10 l 501 truncated Both show the gene pairings and implication index and the components even though the model file simply shows the ID This is used by the program performing the snp snp model expansion Users can estimate the count of snp snp models to be produced by multiplying the two snp counts for any given gene ge
15. 9180 KEGG 05010 by such a report Each snp 176590882 17701871 ENSG00000139180 KEGG 05010 is listed with it s corre what might be produced sponding genes provided with a link to ensembl The text report is similar except the SNPs are listed together as can be seen in the example below SNPS Genes Groups Genes Groups 2072539 1805488 GRIN2B KEGG 05010 GRIN2B KEGG 05010 2072539 1990310 GRIN2B KEGG 05010 ENSGO0000139180 KEGG 05010 176590882 17701871 ENSGQ0000139180 KEGG 05010 Ritchie Lab Software biofilter Reference 21 Gene Gene Models This isn t a report but a data product which can be passed to applications which have been linked to the biofilter library and have support for gene gene models However when the setting BINARY MODEL ARCHIVE is off this is written in plain text allowing the user to see exactly what models their run produced The first line contains the number of models Each subsequent line contains the 4 columns Gene ID 1 Gene ID 2 Implication Index Disease Independent groups associated with the pairing The Gene IDs are the numerical ID value this is the second column from the genes file This file requires the genes output file in order to be used to fully ex pand gene gene models into snp snp models The file is sorted by implication index such that those models with the highest score will be first LEDE 54991 55016 4 1 21419 75847 91412 4 112
16. B I O FI L T ER Reference Manual rev 1 0 1 biofilter Biological Knowledge Integration Utility http chgr mc vanderbilt edu biofilter Table of Contents Introduction 1 Purpose of this manual 1 Conventions Used 1 Example commands 1 biofilter sample config list associations 1 Program Output 1 Configuration details are listed first in bold left aligned with the rest of the text 1 Common Parameters 1 Integer 2 Float 2 Index 2 max 2 min 2 On Off 2 filename 2 label 2 description 2 Using the Biofilter Application 3 Command Line Arguments 3 Ritchie Lab biofilter Reference biofilter config file S sample config 3 C coverage filename 3 D detailed coverage 3 d disease dependent filename 3 filter by genes filename ALL 3 inject gene information analysis results filename integer integer filename ALL 4 G list groups criteria 4 L list models filename 4 model report filename 5 marker info 5 q quiet 5 report gene coverage filename 5 s snps filename 5 snp report 5 W write models float integer 5 m show models filename 5 P list population ids 6 d disease dependent filename 6 p print count estimates 6 strip optimization 6 optimize 6 General Parameters 7 Ritchie Lab biofilter Reference The following parameters control the basic behavior of the application through configuration op tions VARIATION_FILENAME filename
17. E GROUP FILE group list txt This functions identically to INCLUDE GROUPS however the source is a text file This allows users to include a very large number of specially selected groups and their children MODEL FILENAME filename MDOEL FILENAME affy50 models bin This sets the output filename for the binary model file DISEASE DEPENDENT string DISEEASE DEPENDENT alz bio knowledge txt This instructs biofilter to load the knowledge associated with the file alz bio knowledge txt A description of the for mat of these files can be found here POPULATION string POPULATION NO LD Setting the population for a given run allows the user to tap into expanded region boundaries associated with a par ticular LD cutoff and a given population biofilter comes with a set of populations based on LD data found in hap map but users can contribute their own data The string used as the parameter is one of a set of values known to the application Users can query the application for a list of the valid settings by using the list populations flag MODEL_BUFFER_INIT integer MODEL BUFFER INIT 10000 The model buffer is an internal data structure which is used to maintain the massive amount of models generated without requiring huge amounts of RAM This number is used to determine how large the buffer is at start The buffer object will always have at least MODEL BUFFER INIT models in memory at one time and can grow as large as MODEL BUFFER
18. IATION REPORT YES 10 Input File Formats 11 Preferred Aliases 11 SNPS_SOURCE 11 Disease Dependent Genes 11 Disease Dependent Definition 12 Disease Dependent File Format 12 Ritchie Lab biofilter Reference Model Production Overview Reports Coverage Report biofilter sample config s Illumina 660Quad txt report gene coverag alz genes txt C Illumina 660Quad txt Model Summary Report Assocations List biofilter sample config list associations Disease Dependent Contents Report SNP Report biofilter sample config snp report SNP SNP Model Report Gene Gene Models Genes Report SNP Cleanup Report Output Control TBD Example Run Alzheimer s Listing Options biofilter Generating Sample Configuration biofilter sample config gt sample config Target Platform Gene Aliases 14 14 15 15 15 17 18 18 19 20 20 21 22 23 24 25 26 26 26 27 27 28 29 Ritchie Lab biofilter Reference biofilter sample config G alz Report Prefix biofilter list associations Population Selection and LD biofilter sample config P Defining Disease Dependent Groups Model Generation biofilter sample config W 2 0 1000000 tutorial gene gene References Ritchie Lab 29 30 31 32 32 33 34 34 36 38 biofilter Reference Introduction Purpose of this manual Contained within this manual are details for confi
19. MAX MODEL BUFFER INIT during the processing For systems with 1 gigabyte of RAM or less should use the default values Increasing the init size will simply reduce the number of disk reads and thus speed up the processing at the cost of increasing the actual memory foot print during processing It is recommended that the init size be 1 4 the size of the max or smaller MODEL BUFFER MAX integer MODEL BUFFER MAX 100000 This number represents the size the buffer can get before triggering a disk write Larger values will increase perform ance and could reduce the number of disk reads by reducing the number of cycles required On 32bit Redhat intel systems setting the MODEL BUFFER MAX to 20 000 000 allowed very fast processing of 500K sources with a memory footprint of 2 5 Gigabytes The same setting for a million SNP source took 30 minutes and required 2 8 Gigabytes Adding even more memory to the 64bit brought the run down to about 17 minutes Ritchie Lab Software biofilter Reference 500 000 000 MAX_BUFFER The same 1 million SNP source took over 2 hours at 4 000 000 but the footprint was around 1 gigabyte MAX_GENE_COUNT integer MAX GENE COUNT 30 During model production the biofilter ignores groups that exceed this value This is done to avoid generating too many models which defeats the purpose to some extent PREFERRED ALIAS filename PREFERRED ALIAS my favoriate genes txt The user specifies a list of gene aliases
20. at the group is used the variable above can be set high enough to catch the group Users should make these types of changes very carefully though Setting the threshold too high could re sult in massive delays In order to maintain large lists the bilfilter uses a file cache to keep up with the models as they are generated As this cache becomes really large it can alter the runtimes dramatically Population Selection and LD The database that holds group information also contains information about regions genes Users can select a popu lation and an LD threshold to extend gene boundaries during model generation This allows the system to potentially capture SNPs that might carry a signal but are missing from the platform This LD information is drawn from the hapmap project To get a list of populations and their LD thresholds use the list populations P biofilter sample config P Label Comment NO LDNo LD YRI RS1 YRI Population RSquared cutoff of 1 00 YRI RS 8 YRI Population RSquared cutoff of 0 80 YRI RS 7 YRI Population RSquared cutoff of 0 70 YRI DP1 YRI Population DPrime cutoff of 1 00 YRI DP 8 YRI Population DPrime cutoff of 0 80 YRI DP 7 YRI Population DPrime cutoff of 0 70 CEU RS1 CEU Population RSquared cutoff of 1 00 CEU RS 8 CEU Population RSquared cutoff of 0 80 CEU RS 7 CEU Population RSquared cutoff of 0 70 CEU DP1 CEU Population DPrime cutoff of 1 00 CEU DP 8 CEU P
21. be either a fully qualified path such as home torstees wga or it can specified as a path relative to the directory where the applica tion was run such as data goodfilename It can also be just a plain filename as long as the file itself is available from the directory in which the application was run label A label refers to a parameter whose value can be any text string without whitespace These labels are generally used for reporting but in many cases are used to determine filenames As a result users should avoid using unusual char acters in the string that could possibly cause problems with filenames Because spaces and tabs are used to separate each parameter on a given line labels can not contain spaces description A description is a chunk of text that can contain spaces It will always be at the very end of a line and is generally optional Ritchie Lab Software biofilter Reference Using the Biofilter Application The biofilter stand alone application can be used for more than one purpose As a result execution may take more than one set of parameters Many configuration parameters can be overridden on the command line Those parameters are prepended with a wn specially designated keyword which is prefixed by a sign These parameters might take more than one value Be low is a list of the parameters currently supported by the biofilter application All parameters except the configura tion f
22. bio settings cn I1LLumina 660Quad txt variations bn The first three lines describe the source versions used in the local database These refer to the versions from which the data was captured The remaining lines represent the various configuration options in use For the purposes of our example let s make a few changes The following represent some changes that we might make Target Platform Whenever we generate models we need to tell the biofilter what SNPs exist on that platform Users can specify a platform using either configuration changes or on the command line For this example we ll make a single change to the configuration file Open the file and edit the line that contains SNPS_SOURCE Remove the sign and add an appropriate filename after the command It should look something like the following The source file for the RS numbers in your dataset SNPS SOURCE Illumina 66 Quad txt Ritchie Lab Software biofilter Reference 28 The file Ilumina 660Quad txt is just a list of RS Numbers without the letters RS in a simple ascii file that represent each of the SNPs on our platform In this case the list was extracted from documentation on Illumina s 660 Quad platform When specified as we have along with no path information the application assumes that the file is located in the same directory as the application is run Gene Aliases In general the biofilter will supply the gene s
23. e conducting any statistical analysis Rather than annotating the independent effect of each SNP in a GWAS dataset the Biofilter allows the explicit detection and modeling of interactions between a set of SNPs In this manner the Biofilter process provides a tool to discover significant multi SNP models with non significant main effects that have established biological plausibility This approach has the added benefit of reducing both the computational and statistical burden of exhaustively evaluating all possible multi SNP models Model production is gene centric and thus requires that any SNPs to be considered be mapped to genes The gene mapping takes place internally using local copies of current data sources such as Ensembl HapMap and dbSNP A structured mapping is made based on relationships from one of the knowledge sources and this information is used to identify candidates for snp snp models The biological knowledge used by the Biofilter is derived from various sources which are identified as Meta Groups as well as optional user defined groupings Currently the data sources represented include Gene Ontology KEGG Net Path pfam Reactome and DIP These sources are updated periodically and made available as updates at the biofilter website There are two distinct types of data sources Disease dependent sources are user defined and reflect gene disease associations Disease independent sources represent key relationships between genes
24. els Ritchie Lab Software biofilter Reference 18 Disease Dependent Contents Report Suffixed with dd contents this report is produced when the user includes disease dependent information This report is provided to allow the user to verify that the genes being used are the same as the genes were intended Ensembl Start Stop Snp Alias ID ID Chrom Pos Pos Count AGT ENSGQ0000135744 98998 APH1A ENSGQ0000117362 96653 APOA1BP ENSGQ0000163382 97981 154820731 154863290 5 APOA2 ENSGQ0000158874 98229 159458707 159489274 12 1 228777551 228945111 78 1 dl 1 CAMK1G ENSGQ0000008118 98856 1 207793089 207874438 22 1 1 1 148432473 148515725 11 CFH ENSGQ0000000971 98642 194718611 195171294 82 CHRNB2 ENSGQ0000160716 97575 152806881 152818975 1 CLCNKB ENSGQ0000184908 78893 16240720 16272971 18 Ritchie Lab Software biofilter Reference 19 SNP Report SNP reports provide information about the genes for which they are associated If the user has selected LOAD_ALL_ALIASES YES in their configuration a list of corresponding aliases will also be provided This report is available as plain text and HTML format In the HTML format the RS Numbers and gene IDs will appear as links to the corresponding page at the ensembl website As with any other HTML document users should be aware of the potential size of their report before choosing HTML due to potential memory problems biofilter sample config snp report 4 PyNeurGen
25. gain Find the line that starts with POPULATION and change NO LD to CEU DP0 80 Set the population ID to match the population your data is drawn from so that LD patterns can be used to expand the gene boundaries POPULATION CEU DP 80 It is important to realize that the populations listed are the only ones available in the database that is in use Users can process their own LD using thresholds of their own choosing However that is beyond the scope of this tutorial Defining Disease Dependent Groups When building a model list users have the option of adding in their own knowledge into the system This is done through the use of Disease Dependent groups A disease dependent group functions similarly to one of the larger disease independent ones such as KEGG and GO To create a small example disease dependent group create a text file named alz txt containing the following lines ALZHEIMERS GROUP alz assoc Genes previously recognized through association studies AGT APH1A APOA1BP APOA2 CAMK1G CFH CHRNB2 CLCNKB The first line is the name of the meta group name and is used to identify the various groups associated with this dis ease dependent set The next line defines an actual group Each group definition must start with the keyword GROUP followed by it s name and some descriptive commentary Each group should have a unique name since it will be used in the reporting As in all other cases names must co
26. guring and running the application biofilter If this is your first time to use the software we highly recommend that you take a few minutes to download and work through one or more tutorials Then once familiar with the capabilities of the software users can refer to this guide when making changes to the basic configuration settings Conventions Used There are two conventions used throughout this document These text conventions are intended to help distinguish examples from configuration parameters Example commands biofilter sample config list associations The application itself will be named differently according to the platform it was compiled for On Linux systems it will be biofilter or biofilter64 depending on whether it runs on 32bit or 64bit systems Windows and OS X will have be named slightly differently according to their platform as well Program Output Program output will be in a gray box Some lines will often be removed when the output is Lengthy Configuration details are listed first in bold left aligned with the rest of the text The first word s are the keywords which specify what is being changed Each keyword or phrase has some number of parameters These are listed in the order they should appear in the configuration line In some cases parameters can be repeated or are optional Those are denoted inside s Configuration details are generally followed immediately by an example line This i
27. ile are optional Command Line Arguments biofilter config file S sample config Config file Specifies the filename to be used to control genomeSIMLAs overall specific behavior If the configuration is available from within the current working directory the filename alone is sufficient If the filename exists in another directory a fully qualified or relative path should be provided along with the filename itself sample config optional This parameter takes no arguments and causes the output of the biofilter to generate a basic configuration based on the default settings and any that have been overridden by other parameters No other execution is performed C coverage filename filename Adds a file to the coverage report list Coverage files contain RS IDs for a platform used for a coverage re port D detailed coverage Causes the coverage report to provide more detailed information d disease dependent filename filename Adds a disease dependent group to the system This is the same as adding a DISEASE DEPENDENT line to the configuration file filter by genes filename ALL Ritchie Lab Software biofilter Reference filename File containing the genes to be used as a filter Each gene should be on a separate line Generates a comma separated file containing the chromosome in the first column and the RS ID in the second followed by each gene contained within that region in subsequent columns On
28. in important biological proc esses such as a metabolic pathway Bush 2009 Users provide a set of SNPs that reflect the platform on which their analysis will be run This can be a GWAS plat form such as Illumina Human 1M DUo BeadChip or one designed for the user s specific study Biofilter requires only the rsNumber It uses it s own copy of SNP data to insure that positional information matches it s internal representa tion for gene mapping As a result only those SNPs available in the Biofilter s local copy will be considered Ritchie Lab Software biofilter Reference 14 Reports Coverage Report The coverage report shows the amount of coverage a set of genes has on one or more platforms The report requires 2 pieces of information A single gene list and 1 or more coverage lists De pending on the settings chosen the report might be plain text or in HTML agr format As with any other HTML formatted report users should be aware ener of the potential size of the report they are generating since a huge report ace will be difficult to open in most browsers CFH CHRNB2 CLCNKB Gene List The gene list is a series of gene aliases listed on individ Exampte tene ist ual lines These gene aliases must be UniProt IDs or entrez genes and must not be defined such that they represent more than one ensembl ID Coverage Files The user should provide one or more coverage files which are just the individual RS IDs that
29. itchie Lab Software biofilter Reference 30 Assocations List Users can visualize the layout of the grouping information by generating the Association Report This is a hierarchi cal view of the data associated with the requested groups genes and SNPs that will make up the requested model summary The output of the associations list is written to a file ending with associations txt or associations html biofilter list associations Associations Gene Ontonology GO 0048154 1 S100B rs2839357 rs2839362 GO 0048155 1 ATP2A2 rs3026445 rs3026457 rs1860561 GO 0048156 1 S100B rs2839357 rs2839362 G0 0042982 2 G0 0042987 4 ABCG1 rs4148083 rs4148084 rs4148085 rs9975740 rs4148087 rs1117640 rs4148088 rs4148089 rs41480 PSEN1 rs214273 rs8006497 rs36235 rs214260 rs165933 rs362377 ENSGQ0000162736 rs10494342 rs16831846 rs12239946 rs6664438 rs6677637 ENSG 9000143801 rs2073489 rs1295649 ENSG 09000167755 rs1654537 truncated Associations CKEGG KEGG 5010 127 ENSG 09000015475 rs181390 rs181396 rs181402 rs181405 rs9604787 rs181408 rs181417 rs5746474 rs5747351 rs9605401 rs738095 ATP2A2 rs3026445 rs3026457 rs186 561 PSEN1 rs214273 rs8006497 rs36235 rs214260 rs165933 rs362377 NCSTN rs10494342 rs16831846 rs12239946 rs6664438 rs6677637 PSEN2 rs2073489 rs1295640 BACE1 rs7083 rs522843 rs687740 rs47321 rs551662 rs676134 ENSG 9000132906 rs6685648 rs2020902 rs4646018
30. l Minimum Implication Index Set this value to the value of the minimum implication index you wish to use integer optiona Number of models Set this value to the number of models you wish to examine This number takes prece dence over the minimum implication index i e if there are more models with implication index of 3 or greater and the minimum implication index is 2 0 you will still only get models with an implication index of 3 or 4 m show models filename filename Filename indicates the name of the model file file produced during write models command This produces the SNP SNP Model report Ritchie Lab Software biofilter Reference P list population ids Lists the populations available in the database in use Populations are used to adjust the gene boundaries to include additional SNPs that are observed to be within an LD threshold By default there should be 4 popu lations with 3 to 6 different LD thresholds each d disease dependent filename filename This is a specially formatted file which contains user defined group information associated to disease spe cific knowledge This is a command line version of the configuration parameter Disease Dependent Groups A single run can have 0 or more disease specific files p print count estimates This flag forces the application to produce a Gene Gene model report strip optimization This command causes the application to drop all of
31. l C filename options Below is the same report in HTML format using detailed coverage Ensembl IDs and RS Numbers provide links to ensembl s website for convenience Gene Ensembl ID Chromosome Begin kB End kB Hlumina 660Quad txt 593544 18740490 18740492 rs1 1122545 y AGT ENSG00000135744 1 228777551 228945111 78 1811122547 18853459 13200584 1 183118124 rs11588837 rs1 1205 APHIA ENSG00000117362 148432473 148515725 11 rs12086155 rs11205347 rs10888 APOA1BP ENSG00000163382 154820731 154863290 3 184661188 rs11264505 rs120234 186413453 184233368 1s4073054 ENSG00000158874 159458707 159489274 152307425 182501873 rs6686001 1810746420 rs7525534 rs1 1119 ENSG00000008118 207793089 207874438 1811587591 137528698 1366905 1812065463 186428342 1313399 CFH ENSG00000000971 194718611 195171294 1310754192 1310494744 154294 CHRNB2 ENSG00000160716 152806881 152818975 183811450 1810803410 1s6683445 rs868950 CLCNKB ENSG00000184908 16240720 16272971 1312015135 1310803414 rs1204 Boundaries for genes are based on CEU Population DPrime cutoff of 0 80 Ritchie Lab Software biofilter Reference 16 Model Summary Report During model generation a report is made containing all gene gene pairs that were used to generate SNP SNP mod els This report contains SNP counts for each gene the groups in common between the two genes and the number of models produced by
32. ly SNPs that are found in one or more of the genes in filename will appear in the file A new report is generated It s name is based on the REPORT PREFIX configuration with the last portion of the filename being snp report csv inject gene information analysis results filename integer integer filename ALL filename This is the file containing comma separated values for each SNP in the analysis integer This is the column in which the chromosome is found this should start with 1 The chromo some should be 1 22 X Y MT integer This is the column in which the rsid is found the index should start with 1 filename This is the file containing genes which are of interest If the user wants to consider all genes she can choose ALL as the filename When the biofilter performs injection it actually parses the specified CSV file and attempts to append any gene information associated with each SNP in the file The genes that will be identified will be chosen from the second filename parameter and will use the gene boundary extension conventions described elsewhere either using LD Spline for a given population or using a predetermined constant extension up and downstream of the real gene boundaries A new report is generated It s name is based on the REPORT PREFIX configuration with the last portion of the filename being snp analysis csv G list groups criteria criteria optional This string can be used t
33. n To get a list of groups that contain the let ters alz in the name or description field type the following biofilter sample config G alz Meta GroupGroup IDNameDescription Gene Ontonology53887 GO 0048154 Interacting selectively with S100 beta protein S100 is a small calcium and zinc binding protein produced in astrocytes that is implicated in Alzheimer s disease Down Syndrome and ALS GOC jic Gene Ontonology53892 GO 0048155 Interacting selectively with S100 alpha protein S100 is a small calcium and zinc binding protein produced in astrocytes that is implicated in Alzheimer s disease Down Syndrome and ALS GOC jic Gene Ontonology 53894 GO 0048156 Interacting selectively with tau protein tau is a microtubule associated protein implicated in Alzheimer s disease Down Syndrome and ALS GOC jic Gene Ontonology 101057 G0 0042982 The chemical reactions and pathways involving amyloid precursor protein APP the precursor of beta amyloid a glycoprotein associated with Alzheimer s disease GOC go_curators Gene Ontonology 101064 GO 0042987 The chemical reactions and pathways resulting in the breakdown of amyloid precursor protein APP the precursor of beta amyloid a glycoprotein associated with Alzheimer s disease GOC go_curators Gene Ontonology106734G0 0050435 The chemical reactions and pathways involving beta amyloid a glycoprotein associated with Alzheimer s disease and its precursor amyloid precursor p
34. n this example The examples listed below use the following command to execute the biofilter biofilter This command may differ from machine to machine depending on hardware This is done to allow different versions of the software to exist side by side In general 32bit linux distributions will simply be called biofilter When compiled for 64bit systems the name will be biofilter64 For OSX and windows the name will be biofilter OSX and biofilter win32 with the possible 64 following the word biofilter when built for 64bit Listing Options If you are completely new to the biofilter a good first step is to take a quick look at the various options available As with many Unix application simply running the program with no parameters will generate a basic list of options S biofilter biofilter 0 5 0 600 Debug Tue Sep 1 16 07 51 CDT 2009 Marylyn Ritchie William Bush and Eric Torstenson Please forward any comments or errors to biofilter chgr mc vanderbilt edu usage biofilter lt configuration file gt biofilter is a standalone application for use in investigating possible SNP associations in a set of data which through biological knowledge might be worth investigating Optional Commands Include S sample config Print sample configuration to std out report gene coverage gene list filename Reports the snp count for the genes in genelist for the snps in snp source marker info Reports each SNP and it s positi
35. ne model Ritchie Lab Software biofilter Reference 36 Finally there is the file tutorial snpsnp This contains the actual snp snp models of interest and may be binary for the same reason as the gene gene model file 1004632 230 230 230 230 230 230 230 Lo 13490 13728 14210 14576 27152 27154 27827 NNNNNNN truncated The first line is simply the number of models found in the file Each subsequent line lists two RS IDs without the letters R and S and the Implication Index This file is sorted by rs ID 1 and rs ID2 respectively and should con tain no duplicate rs pairings Another file can be found called tutorial genes This file can be used by programs reading the gene gene models and describes the SNPs associated with a particular gene gene model production Ritchie Lab Software biofilter Reference 37 References Bush WS Dudek SM Ritchie MD Biofilter A Knowledge Integration System for The Multi locus Analysis of Genome wide Association Studies Pacific Symposium on Biocomputing 2009 368 379 Ritchie Lab Software biofilter Reference 38
36. never complete MODEL_BUFFER_MAX 100000 Set the population ID to match the population your data is drawn from so that LD patterns can be used to expand the gene boundaries POPULATION NO LD Add one or more files containing disease dependent genes DISEASE_DEPENDENT User can specify aliases for genes the alias must be present in the database PREFERRED_ALIAS Prefix used for all reports REPORT_PREFIX Loads all aliases and generates a text report containing their associations LOAD_ALL_ALIASES NO sample config Many configuration options also have a command line override available This is useful for performing tasks which are repeated from time to time Ritchie Lab Software biofilter Reference 27 General Reporting Except when generating a sample configuration a summary report is produced containing the configuration details being used along with some details about the database in use SESE SEIS Dependency Versions dbSNP 36 Ensembl 27 Hap Map LD 53 a eo Configuration Parameters DISEASE_DEPENDENT alz txt INCLUDE_GROUPS REPORT_PREFIX SETTINGS_DB SNPS SOURCE VARIATION FILENAME 53887 53892 53894 101057 101064 106734 160871 LOAD_ALL_ALIASES NO MAX_GENE_COUNT 30 MODEL_BUFFER_INIT 10000 MODEL_BUFFER_MAX 100000 MODEL_FILENAME NONE POPULATION CEU DP 80 PREFERRED_ALIAS gene_aliases txt PROJECT sample config tutorial
37. ntain no spaces However the description can con tain any character the user prefers except for new line characters Following the group definition are the gene names These should be common names and must appear in the list of known aliases In general those names should be recognized by EntrezGene or Uniprot and must identify a single gene Gene aliases should be separated by whitespace however they must start on the line after the group definition We could add more groups to this file using addition GROUP definitions but this is sufficient for our needs Adding the disease dependent information to a run can be done using either a configuration file or on the command line We ll add it to the configuration file here Open the configuration and add the new filename to the end of the line starting with DISEASE_ DEPENDENT Be sure to remove that as well or else the line will be ignored Add one or more files containing disease dependent genes DISEASE_DEPENDENT alz txt Ritchie Lab Software biofilter Reference 33 Model Generation We are now ready to produce our model list The most important step is the production of a gene gene model list This is a complete listing of all gene gene models that were defined based on the Disease Independent information in our database and the disease specific information provided by the user This model list can be passed directly to some programs such as the application athena
38. o search the group s name and description If a match is found the Meta Group name group ID group name and group description are displayed for each match If this command appears as the last flag on the command line all groups will be listed This list does not refer to any setting other than the value at SETTINGS_DB As a result all groups that are part of the SETTINGS_DB are considered regardless of restrictive group limitations set within the configura tion L list models filename filename The name of the model file from a previous run Ritchie Lab Software biofilter Reference This produces a basic model report model report filename filename File containing pairwise snp models Each line contains a single pairwise model listed as integer values separated by whitespace marker info Produce haploview compatible marker info files based on the SNPS SOURCE platform q quiet Turns off all non vital output to stdout and stderr report gene coverage filename filename This file holds the list of genes for which coverage is to be reported in the coverage report s snps filename filename This overloads the setting SNPS_SOURCE from the configuration file snp report Produces a report containing all genes associated with each of the SNPs found in SNPS_SOURCE W write models float integer Produces the gene gene models and optionally produces SNP SNP models as well float optiona
39. on chromosome in a format acceptable by haploview snp report For each SNP in the SNP Source lists the genes where that RS number is found Llist associations Lists the associations for each group Optional Parameters Include s snps lt snps filename gt Qverride the snp source file ont he commandline C coverage lt snps filename gt Add a file to coverage report list D detailed coverage used with C adds extra details to coverage report L list models Writes model list to std out W write models lt model filename gt Writes model list to file in binary format m show models lt model filename gt Writes contents of model file to screen in human readable form p print count estimates Lists count estimates for gene gene models 1 load ld lt model filename gt Loads LD information from the file filename and adjusts the gene boundaries accordingly d disease dependent lt filename gt Adds a meta group containing data from the file filename G list groups criteria Adds group search criteria and produces a list of group IDs that match the criteria P list populations Lists all available Population based LD boundary options optimize Updates internal structures to allow faster access This is usually done prior to release strip optimization Strips the optimization out this is helpful to allow data imports to run more quickly
40. opulation DPrime cutoff of 0 80 CEU DP 7 CEU Population DPrime cutoff of 0 70 CHB RS1 CHB Population RSquared cutoff of 1 00 CHB RS 8 CHB Population RSquared cutoff of 0 80 CHB RS 7 CHB Population RSquared cutoff of 0 70 CHB DP1 CHB Population DPrime cutoff of 1 00 CHB DP 8 CHB Population DPrime cutoff of 0 80 CHB DP 7 CHB Population DPrime cutoff of 0 70 JPT RS1 JPT Population RSquared cutoff of 1 00 JPT RSQ 8Q0JPT Population RSquared cutoff of 0 80 JPT RSQ 70JPT Population RSquared cutoff of 0 70 JPT DP1 JPT Population DPrime cutoff of 1 00 JPT DP 8 JPT Population DPrime cutoff of 0 80 JPT DP 7 JPT Population DPrime cutoff of 0 70 By default the system uses NO LD which interprets a gene explicitly by it s beginning and end base pair location The procedure used to extend the boundaries is called LD Spline a technique developed by the Ritchie Lab The cut off mentioned in the report above is the minimum LD statistic R Squared or D a pair can have to extend the re gion s boundaries Ritchie Lab Software biofilter Reference Our study most closely matches the CEU population so we ll use one of those options We ll let the system be quite liberal in the definition of a region s boundaries so we ll choose the middle D option 0 80 To indicate to the biofilter that a particular population and LD threshold is to be used the configuration file must be edited once a
41. oviding groups of genes that have been known to be related to a given disease it is possible to add models that represent interactions that a highly relevant to the disease that might not exist in more traditional grouping paradigms Disease Dependent Definition Users can create many disease dependent groupings A high level grouping or meta group will contain one or more group Each of these groups will contain one or more regions At present the biofilter only looks at genes within a group for constituent models and there is no support for hierarchical groupings However users can simulate the ef fects of hierarchy by creating different disease dependent meta groups and correctly choosing the one appropriate for their needs Models found inside multiple groups within a single meta group will have their implication index incre mented only by one If a model occurs inside multiple meta groups it s implication index will reflect each meta group it is found inside Disease Dependent File Format Disease dependent configurations are done outside of the application using a simple text file format Each file repre sents a super group meta group which can contain one or more groups Name Description The first line of the file must contain the disease dependent name a string with no spaces followed by a short de scription this can have spaces but must fit on a single line The total length of the line should be 4096 character
42. owever in the event that major changes have been identified the file require a newer version of the applica tion In this event users will be notified and will have to update their software to use the newest data MAX GENE COUNT integer MAX GENE COUNT 30 This sets the upper limit for acceptable group size Acceptable group size is the number of genes contained by a given group If a group has as many or fewer genes in it the biofilter uses it s contents to generate models If a group ex ceeds this number the biofilter queries each of it s children groups performing the same check This setting is used to constrain the size of the resulting model counts to a number that is manageable according to modern computation resources SNPS SOURCE filename SNPS SOURCE projects ritchie biofilter Affy6 0 v27 txt This setting allows the user to limit the SNPs considered to a specific platform or some user defined set of SNPs By setting this value the biofilter only loads SNP data for those found in the source INCLUDE GROUPS group group group This allows the user to limit the search to those groups and their children This can be a specific group within a meta group or it can be a meta group ID such as DIP Each group is separated by spaces and is the group s unique ID When including a group all of that groups children are also included Ritchie Lab Software biofilter Reference INCLUDE GROUP FILE filename INCLUD
43. rom the SNPS_SOURCE file the biofilter attempts to identify any RS IDs that have been merged into previous RS IDs by dbSNP Any RS IDs that have been merged will be replaced with the proper RS ID Those SNPs that have been identified as having been deleted by NCBI are removed Each removal and renaming is noted in the file projectname snp cleanup This file is a tab separated file listing first the SNPs that were removed due to having been deleted at NCBI followed by the mapping details for those that have been merged 4 Expired SNPs Encountered 45469397 rs45552437 rs45616434 rs45628831 504 rs IDs were updated Original ID rs41429248 rs41480744 rs12782608 rs41396045 rs41495349 rs4365706 rs11242845 rs41503946 rs4446752 Ritchie Lab Software New ID rs2305130 rs17145687 rs9422653 rs36130286 rs16865746 rs4026962 rs9501985 rs11077998 rs2314691 truncated 24 biofilter Reference Output Control TBD Ritchie Lab Software biofilter Reference Example Run Alzheimer s The following tutorial will walk the user through performing many typical biofilter tasks including preparing a highly specialized model set to use with plato for performing an association study using knowledge specific to Alz heimer s disease This search is intentionally highly selective and is intended for instructional purposes only Under most circumstances users will want to incorporate more knowledge than we will be including i
44. rotein APP GOC ai KEGG160871KEGG 5010Alzheimer s disease Ritchie Lab Software biofilter Reference 29 Users can perform more selective searches by adding additional G keyword phrases to the command line The more keywords you add the more selective the search will be The second column contains the group IDs which is what we need to update our configuration file so that it only includes the groups we are interested in Open the configuration file and change the line containing IN CLUDE GROUPS so that it looks similar to the text below List the various groups by group name separated by spaces INCLUDE_GROUPS 53887 53892 53894 101057 101064 106734 160871 Notice that groups are separated by spaces not commas When users specify no groups all groups are used However if one or more groups are added to this line only those groups and any that are hierarchically contained within them are used to generate models Users can include an en tire group by using the group ID associated with the top level group such as the ID for GO Report Prefix One last detail is the report prefix By default any text output that doesn t go to the command line will be named after the configuration file However users can override this behavior by providing a Report Prefix For our purposes we ll simply change the reports to start with the word tutorial Prefix used for all reports REPORT_PREFIX tutorial R
45. s an example Examples show how an actual entry would look and are followed by some descriptive information to help the user understand how the example would affect the biofilter application runtime Common Parameters Ritchie Lab Software biofilter Reference There are a number of parameters which are used commonly across multiple configuration settings In order to sim plify the descriptions of the various properties of each command we ll describe those properties here and just refer to them as if they were a type Integer Parameters specified in this way just simply refer to a whole number In general these values should be equal to or greater than 0 except when specified otherwise Float Values specified as float are decimal values Index If a parameter is listed as an index it refers to the index starting at 1 the user wishes to select max This is generally an integer value representing the upper bound of some value In some cases such as minor allele frequency it might represent a floating point value min This is generally an integer value representing the lower bound of some value In some cases such as minor allele frequency it is possible that it represents a floating point value On Off These parameters accept a boolean Yes No type setting Users can use ON OFF or YES NO to set them filename When a configuration refers to a file for input or output the filename is generally used This can
46. s or less These are used by for reporting purposes and should be as meaningful as possible i e if the user will have more than one meta group for a single disease they should properly name them so that they can be distinguished from one another GROUP group name group description GROUP This keyword is required and should be correctly capitalized Group Name String with no spaces or tabs inside This is used for reporting purposes and should be meaningfully distinguishable from any other groups Group Description Simple description for the group This can have spaces but must fit on the line with a total length of 4096 or less gene alias gene alias Ritchie Lab Software biofilter Reference Gene aliases are gene identifiers from one of the following sources Uni Prot IDs TREMBL or SwissProt Entrez Gene These aliases must only identify a single entity have no spaces in them and must be available from within the Ensembl build upon which the biofilter data was based Aliases can appear on separate lines or with spaces tabs separating them or a mix of the two ALZHEIMERS GROUP alz assoc Genes previously recognized through association studies AGT APH1A APOA1BP APOA2 CAMK1G CFH CHRNB2 CLCNKB Ritchie Lab Software biofilter Reference 13 Model Production Overview The Biofilter uses biological information about gene gene relationships and gene disease relationships to construct multi SNP models befor
47. ssociation with disease dependent genes When set to ALL MOD ELS default all models will be produced When set to GROUP LEVEL only groups where gene gene models are Ritchie Lab Software biofilter Reference produced that has one or more genes found in a disease dependent group will yield models When set to DD_ONLY only the gene gene models in which one of the genes is found in one of the disease dependent groups are produced COLLAPSE_ASSOCIATION_REPORT YES NO COLLAPSE ASSOCIATION REPORT YES When set to yes the associations shown reflect only the groups which could generate models The report properly respects the MAX_GENE_COUNT setting but it does not obey the DISEASE_DEPENDENT_LEVEL setting it shows all groupings despite their relationship with disease dependent groups BINARY_MODEL_ARCHIVE YES NO BINARY MODEL ARCHIVE YES Setting this value to YES causes the gene gene model archive and the snp snp model archive to be written in binary format In most cases this won t be necessary since both files will probably be reasonable in length However if ei ther are expected to contain more than a few hundred thousand entries it is recommended to use ASSOCIATION REPORT YES NO ASSOCIATION REPORT YES When Yes this causes the biofilter to produce an association report Ritchie Lab Software biofilter Reference 10 Input File Formats With very few exceptions files are space delimited ascii files Preferred
48. the Indexes This can be done to speed up the insertions necessary during LD imports Once the data has been properly imported users should be sure to optimize the data once again optimize This adds indexes to the underlying database if they don t exist This is generally only done once but can be removed prior to performing LD import see strip optimization above Users generally shouldn t need to use this command Ritchie Lab Software biofilter Reference General Parameters The following parameters control the basic behavior of the application through configuration options VARIATION_FILENAME filename VARIATION_FILENAME variations bn This sets the path to the source file containing the binary variation data This file should originally be downloaded with the application however as the data is updated this file can be downloaded independently of the database There is the possibility that the file is incompatible with the version of the biofilter in use In such cases the applica tion will exit with an appropriate measure To resolve this issue the user should download the latest version of the application as well SETTINGS DB filename SETTINGS DB bio settings cn This sets the path to the settings database This file is required for all biofilter functionality and will periodically up dated with new information Users generally will be able to update their settings file without updating the applica tion h
49. the pairing By default Genes are reported using their Ensembl Stable ID However users can configure a Preferred Gene Alias file Genes which appear in the preferred alias file will be reported according to the preferred alias SN Name Count Name Count Index Count DI DD ABCG1 7 ACVR2B 7 2 49 1 485522 ABCG1 7 ABCB1 58 2 406 1 485522 ABCG1 rd CYP3A4 37 2 259 1 485522 ABCG1 7 SLC22A11 7 2 49 1 485522 ABCG1 7 SCUBE3 47 2 329 1 485522 ABCG1 7 ANXA9 6 3 42 112 485523 ABCG1 7 SULF1 3 2 21 1 485522 ABCG1 7 STX4 17 2 119 1 485522 ACVR2B 7 ABCB1 58 2 406 1 485522 ACVR2B 7 CYP3A4 37 il 259 1 ACVR2B 7 SLC22A11 7 a 49 i ACVR2B 7 SCUBE3 47 329 iL ACVR2B 7 ANXA9 6 2 42 1 485523 SULF1 3 ACVR2B 7 1 21 1 STX4 17 ACVR2B 7 1 119 Ls ABCB1 58 CYP3A4 37 2 2146 1 485522 SLC22A11 7 ABCB1 58 2 406 1 485522 SCUBE3 47 ABCB1 58 2 2726 1 485522 Ctruncated Ritchie Lab Software biofilter Reference 17 Assocations List Suffixed by associations txt or associtions html association reports provide the user with a hierarchical represen tation of groups in their search Only groups that have been loaded using INCLUDE_GROUPS or all if no groups were provided will appear biofilter sample config list associations Associations Gene Ontonology GO 0048154 1 S10 B rs2839357 rs2839362 G0 0048155 1 ATP2A2 rs3026445 rs3026457 rs186 561 GO 0048156 1 S10 B rs2839357 rs2839362 GO 0042982 2 GO 0042987
Download Pdf Manuals
Related Search