SNPsyn 1.2b is a tool for gene interaction analytics.
(C)2011-2014 University of Ljubljana, Bioinformatics Laboratory
and Laboratory for Adaptive Systems and Parallel Processing

Running: man - Display man page.

usage: SNPsyn <command> [options] [arguments]
Type 'SNPsyn help <command>' for detailed help.

Available commands:
  singles   - Calculate information gain of single markers.
  pairs     - Calculate synergy (interaction gain) and information gain of marker pairs.
  display   - Display results (histograms, lists of top markers, etc) stored in a binary file.
  merge     - Merge results (histograms, lists of top markers, etc) from individual chunks.
  normalize - Normalize a distribution (1D or 2D histogram).
  pFDR      - Calculate probability and FDR maps.
  extract   - Extract 1D histograms from 2D histogram.
  addsig    - Add p-value and FDR scores to a list of top singles or top pairs.
  ped2syn   - Convert a data file from .ped (PLINK's format) to .syn (SNPsyn's format).
  tab2syn   - Convert a data file from .tab (tab delimited) to .syn (SNPsyn's format).
  syn2tab   - Convert a data file from .syn (SNPsyn's format) to .tab (tab delimited).
  help      - Display help of commands.
  man       - Display man page.


singles - Calculate information gain of single markers. usage: singles [options] datafile input: datafile - GWAS data file in .syn format output: singles_prob_chunks_chunk.bin or singles_prob_chunks_chunk_rndseed.bin, if random seed (-r) is set to non-zero Where prob is the probability estimation used (lap or rel), chunks is the number of chunks into which the analysis was divided into chunk is the chunk on which the analysis was performed, and rndseed is the given, non-zero random seed. Use option -o to change prefix 'singles'. Valid options: -p (lap|rel) - probability estimation (default: lap). lap - Laplace estimate rel - relative frequency -r integer - use given random seed to permute data prior to analysis (default: 0). Data are not permuted if random seed equals zero. -k integer - report on top results only (default: 5000). Reports on top k results (singles, pairs, etc) only. Report no results if k equals zero. -o output_name - write results to file (default: singles). Use this option to change the predefined file name that is generated by the command. -g double - granularity (precision) of outputed histograms (1D and 2D) (default: 0.0015). Valid values are from 0.001..0.02. -m double - allowed ratio of missing values (default: 0.1). Valid values are from 0.0..0.5. Markers with higher ratio of missing values will not be analysed. -G integer - GPU support (default: 0). 0 - do not use GPU >0 - select GPU device by index.
pairs - Calculate synergy (interaction gain) and information gain of marker pairs. usage: pairs [options] datafile input: datafile - GWAS data file in .syn format output: pairs_prob_chunks_chunk.bin or pairs_prob_chunks_chunk_rndseed.bin, if random seed (-r) is set to non-zero Where prob is the probability estimation used (lap or rel), chunks is the number of chunks into which the analysis was divided into chunk is the chunk on which the analysis was performed, and rndseed is the given, non-zero random seed. Use option -o to change prefix 'pairs'. Valid options: -p (lap|rel) - probability estimation (default: lap). lap - Laplace estimate rel - relative frequency -h (screen|main|off) - two-stage heuristic (default: screen). off - perform exhaustive search (may take very long!) main - First, select a subset of 22000 most informative SNPs, then search all pairs among selected SNPs. screen - Score all pairs with a fast but approximate upper bound on Syn. Select 22000*21999/2 best scored pairs and calculate Syn, I, etc. Both two-stage heuristics will substantially speed-up search, but at the cost of missing some highly synergistic pairs. -c chunks - divide the analysis into chunks (default: 1). This option is used in conjunction with option -i. -i chunk - divide and perform the analysis on a chunk (default: 0). Valid values are from 0..chunks-1. This option is used in conjunction with option -c. -j chunk_stop - divide and perform the analysis on an interval of chunks (default: 0). Valid values are from 0..k, where k=min(i, chunks-1). Lower bound (i) is determined by parameter -i. This option is used in conjunction with options -c and -i. -r integer - use given random seed to permute data prior to analysis (default: 0). Data are not permuted if random seed equals zero. -k integer - report on top results only (default: 5000). Reports on top k results (singles, pairs, etc) only. Report no results if k equals zero. -o output_name - write results to file (default: pairs). Use this option to change the predefined file name that is generated by the command. -g double - granularity (precision) of outputed histograms (1D and 2D) (default: 0.0015). Valid values are from 0.001..0.02. -m double - allowed ratio of missing values (default: 0.1). Valid values are from 0.0..0.5. Markers with higher ratio of missing values will not be analysed. -G integer - GPU support (default: 0). 0 - do not use GPU >0 - select GPU device by index. -M integer - MIC support (default: 0). 0 - do not use MIC 1 - use MIC
display - Display results (histograms, lists of top markers, etc) stored in a binary file. usage: display [options] file.bin input: file.bin - file with results in binary format File type is determined automatically and its content displayed on standard output, in human-readable format. Supported file types are: 1D histograms, 2D histograms, FDR map, probability map, list of single most informative markers, list of most synergistic pairs, list of most informative pairs. Valid options: -k integer - report on top results only (default: 5000). Reports on top k results (singles, pairs, etc) only. Report no results if k equals zero.
merge - Merge results (histograms, lists of top markers, etc) from individual chunks. usage: merge [options] file1.bin file2.bin ... input: file1.bin... - file(s) to merge (in binary format) output: merged.bin - filename where merged results are stored into Use option -o to change the output filename. Files to merge must be of same type and are determined automatically. Supported file types are: 1D histograms, 2D histograms, FDR map, probability map, list of single most informative markers, list of most synergistic pairs, list of most informative pairs. Valid options: -k integer - report on top results only (default: 5000). Reports on top k results (singles, pairs, etc) only. Report no results if k equals zero. -o output_name - write results to file (default: merged.bin). Use this option to change the predefined file name that is generated by the command.
normalize - Normalize a distribution (1D or 2D histogram). usage: normalize [options] file.bin input: file.bin - file to normalize output: normal.bin - file where normalized results will be saved Use option -o to change the output filename. Type of file (1D or 2D histogram) is determined automatically. The histogram is first normalized (to sum 1.0) and then saved into the output file. Valid options: -o output_name - write results to file (default: normal.bin). Use this option to change the predefined file name that is generated by the command.
pFDR - Calculate probability and FDR maps. usage: pFDR [options] true.bin random.bin input: true.bin - histogram of results on true data random.bin - histogram of results on randomly permuted data output: [prefix].probmap.bin - probability map [prefix].FDRmap.bin - FDR map Use option -o to change prefix of the output files. Type of distribution (1D or 2D histograms) is determined automatically. The prefix of the output files can be changed with the -o parameter. Valid options: -o output_name - write results to file (default: map). Use this option to change the predefined file name that is generated by the command.
extract - Extract 1D histograms from 2D histogram. usage: extract [options] histogram2d.bin input: histogram2d.bin - 2D histogram output: [prefix].marginal_i.bin - collapsed 2D histogram onto the x-axis (information gain) [prefix].marginal_syn.bin - collapsed 2D histogram onto the y-axis (synergy) If not given with parameter -o, prefix is the name of the input file. Valid options: -o output_name - write results to file. Use this option to change the predefined file name that is generated by the command.
addsig - Add p-value and FDR scores to a list of top singles or top pairs. usage: addsig [options] toplist.bin probmap.bin FDRmap.bin input: toplist.bin - file with results (singles, pairs) probmap.bin - file with probability map (1D or 2D) FDRmap.bin - file with FDR map (1D or 2D) output: toplist.bin - updated results are saved in same (input) file unless specified otherwise. Use option -o to change the output filename. Type of results (single most informative markers, mort synergistic paris, most informative pairs) is determined automatically. Results are first read from the input file. Each record gets scored according to the given probability and FDR maps. Results (with probability and FDR scores) are then saved into the output file. Valid options: -o output_name - write results to file. Use this option to change the predefined file name that is generated by the command.
ped2syn - Convert a data file from .ped (PLINK's format) to .syn (SNPsyn's format). usage: ped2syn data.ped data.map data.syn input: data.ped - file with genotype data (PLINK's format) data.map - companion file to data.ped (PLINK's format) output: data.syn - file to where save the data (SNPsyn's format) Commands for interaction analysis ('singles' and 'pairs') require data in SNPsyn's format. Valid options: -m double - allowed ratio of missing values (default: 0.1). Valid values are from 0.0..0.5. Markers with higher ratio of missing values will not be analysed.
tab2syn - Convert a data file from .tab (tab delimited) to .syn (SNPsyn's format). usage: tab2syn data.tab data.syn input: data.tab - file with genotype data (tab-delimited format) output: data.syn - file to where save the data (SNPsyn's format) Commands for interaction analysis ('singles' and 'pairs') require data in SNPsyn's format. The .tab file must include columns with data on markers, and also a binary attribute indicating two phenotypes (cases and controls). Name of the attribute must be given with option -ph. Valid options: -m double - allowed ratio of missing values (default: 0.1). Valid values are from 0.0..0.5. Markers with higher ratio of missing values will not be analysed. -ph string - phenotype attribute name. Specify which binary attribute to use for phenotype. Attribute must be binary (representing cases and controls). Can be of form attribute=[case_value], where [case_value] indicates the phenotype value representing cases. If not given, first phenotype value listed is interpreted as case.
syn2tab - Convert a data file from .syn (SNPsyn's format) to .tab (tab delimited). usage: tab2syn data.syn data.tab input: data.syn - file with genotype data (SNPsyn's format) output: data.tab - file to where save the data (tab-delimited format This command is useful to convert the data from SNPsyn's (.syn) format into a human-readable format.
help - Display help of commands. usage: help command Valid options: -H 0,1 - output in HTML format (default: 0).
man - Display man page. usage: man Valid options: -H 0,1 - output in HTML format (default: 0).
See http://snpsyn.biolab.si for more information.