For more information see wiki
PARSES is intended to be executed with Solexa data on a *NIX-based desktop computer. It may not be used for any sort of financial gain. Licenses for both MEGAN and Novoalign are strictly for non-profit research at non-profit institutions and academic usage.
PARSES's system requirements are directly dependent on the size of the data set being processed. It is recommended to be run on a Linux or OS X 64-bit machine with at least 4GBs of memory. You must also have root privileges to the machine if you are installing software.
- Working directory is the directory in which the FASTQ sequence file is contained.
- abyssKmerOptimizer.pl marked as executable in same folder as rakefile
- fac.pl marked as executable in same folder as rakefile
- addTaxon.pl marked as executable in same folder as rakefile
- parallelBlast.sh marked as executable in same folder as rakefile
- Xextractspans.pl marked as executable in same folder as rakefile
- Xfilterspans.pl marked as executable in same folder as rakefile
- Xnovotonm.pl marked as executable in same folder as rakefile ___
Execute all commands from the directory of the data. Place all scripts for PARSES into a single directory and mark as executable. All latest versions of programs will be installed, in the event of an error during installation a repository of the programs may be used which is not guaranteed to be up to date. The repository can be activated by including the repo=true command. Installation will automatically be performed during any execution but it can be manually performed by evoking any of the following installation commands:
sudo rake -f /rake/file/location install #installs and indexes all resources. sudo rake -f /rake/file/location novoalignInstall sudo rake -f /rake/file/location bowtieInstall sudo rake -f /rake/file/location hgInstall sudo rake -f /rake/file/location novoIndex sudo rake -f /rake/file/location bowtieIndex sudo rake -f /rake/file/location samtoolsInstall sudo rake -f /rake/file/location tophatInstall sudo rake -f /rake/file/location abyssInstall sudo rake -f /rake/file/location blastInstall sudo rake -f /rake/file/location ntInstall sudo rake -f /rake/file/location meganInstall sudo rake -f /rake/file/location parallelIteratorInstall
rake -f /rake/file/location seq=NameYouGiveToYourSequence file=YourSequenceFileName.fastq type=illumina1.3
rake -f /rake/file/location seq=NameYouGiveToYourSequence
rake -f /rake/file/location repo=true install
rake -f /rake/file/location seq=NameYouGiveToYourSequence file=YourSequenceFileName.fastq type=illumina1.3 localAlignContigs
rake -f /rake/file/location seq=NameYouGiveToYourSequence file=YourSequenceFileName.fastq type=illumina1.3 truncate=true localAlignContigs
Run truncated version of PARSES (only execute the specified task and ignore prerequisites) and override the file naming schema
rake -f /rake/file/location seq=NameYouGiveToYourSequence file=YourFileToProcess type=illumina1.3 truncate=true forcefile=true localAlignContigs
rake -f /rake/file/location -T
- alignSequence (Novoalign)
- removeHuman (Xnovotonm)
- removeSpans (Tophat)
- localAlignReads (BLAST+)
- metaGenomeAnalyzeReads (MEGAN)
- denovoAssembleCluster (ABySS)
- localAlignContigs (BLAST+)
- metaGenomeAnalyzeContigs (MEGAN)
- alignSequence # Novoalign - Align reads in to base genome
- removeHuman # Xnovotonm - Harvest non-base organism reads
- removeSpans # Tophat - Align spanning reads in order to remove base organism reads
- localAlignReads # BLAST - Associate reads with organisms.
- metaGenomeAnalyzeReads # MEGAN - Separate reads into taxonomies.
- denovoAssembleCluster # ABySS - Assemble reads associated with clusters of taxonomies.
- localAlignContigs # BLAST - Associate contigs with organisms.
- metaGenomeAnalyzeContigs # MEGAN - Separate contigs into taxonomies.
- clean # Remove any temporary products.
- clobber # Remove any generated file.
- reserialize # Automatically saving any settings changes which may have been made
- abyssInstall # Install latest version of ABySS with Google Sparsehash.
- blastInstall # Install latest version of BLAST+.
- bowtieIndex # Create an index for bowtie/tophat of the human genome database.
- bowtieInstall # Install latest version of Bowtie.
- hgInstall # Install latest version of human genome database.
- install # Install latest version of everything.
- meganInstall # Install latest version of MEGAN.
- novoIndex # Create an index for novoalign of the human genome database.
- novoalignInstall # Install latest version of Novoalign.
- ntInstall # Install latest version of the NT database.
- parallelIteratorInstall # Install latest version of Parallel::Iterator for perl.
- samtoolsInstall # Install latest version of Samtools.
- tophatInstall # Install latest version of Tophat.
The primary method for viewing results involves perusing the RMA file produced by MEGAN--though a PDF is also produced with standard LCA parameters. To familiarize yourself with MEGAN you can watch the MEGAN Introduction Youtube video.
Your results will resemble the following:
For a detailed example of what results changing LCA parameters will have click the image below (or here for a PDF version).
Reporting is currently stored in logs.
A log file is produced in the data directory with the filename
sequenceName.log. It not only logs all commands executed, in addition to errors returned, but it also contains report information on the results of the data analysis. Below is a list of all information logged at each step of PARSES.
- Sequence name.
- Read length.
- Data type.
- Timing information for each portion of PARSES.
- Arguments used for each program.
- Information regarding the specifications of the machine on which PARSES is executing. (log)
- Command used to invoke PARSES. (log)
- Total number of reads. (log)
- Number of reads which align to the human genome. (log)
- Percentage of total reads which align to the human genome. (log)
- Number of reads left after removing human reads. (log)
- Percentage of total reads left. (log)
- Number of reads left after removing splicing junctions of human reads. (log)
- Percentage of reads left after removing splicing regions. (log)
- Potential number of exogenous reads. (log)
- Number of reads assigned to a taxonomy. (log)
- Percentage of total reads assigned to a taxonomy. (log)
- Number of reads not assigned to a taxonomy. (log)
- Percentage of total reads not assigned to a taxonomy. (log)
- Number of reads which did not significantly align. (log)
- Percentage of total reads which did not significantly align. (log)
- Image of taxonomy overview with default LCA parameters.
- Coverage threshold. (log)
- Median kmer coverage. (log)
- Number of kmers in contigs. (log)
- Number of contigs. (log)
- Optimum kmer. (log)
- Potential number of exogenous contigs. (log)
- Number of contigs assigned to a taxonomy. (log)
- Number of contigs not assigned to a taxonomy. (log)
- Number of contigs which did not significantly align. (log)
Below is an example of a log file:
# Logfile created on Mon Nov 29 20:21:37 -0600 2010 by logger.rb/22285 I, [2010-12-27T14:54:16.761320 #94508] INFO -- : Begin run for seq=akata file=s_4_sequence_Akata.txt type=illumina1.3 task=-f I, [2010-12-27T14:54:16.849927 #94508] INFO -- : Executing PARSES v0.30 in a 0 environment with 64GB of memory and 24 cores with a 64-bit architecture. I, [2010-12-27T17:56:20.865114 #94508] INFO -- : tophat -p 24 --solexa1.3-quals --output-dir akata_tophat_out /usr/share/hgChrAll s_4_sequence_Akata.txt I, [2010-12-27T17:56:20.865442 #94508] INFO -- : samtools view -h -o akata_tophat_out/accepted_hits.sam akata_tophat_out/accepted_hits.bam I, [2010-12-27T17:56:20.865475 #94508] INFO -- : Xextractspans.pl akata_tophat_out/accepted_hits.sam I, [2010-12-27T17:56:20.865506 #94508] INFO -- : Xfilterspans.pl s_4_sequence_Akata.txt.novo.NM.fasta akata_tophat_out/accepted_hits.sam.spans
A configuration file is produced for PARSES in
$HOME/.PARSES. It contains the paths to the human genome and NT databases as well as the paths to the Bowtie and TopHat indices. It has the following form:
--- !ruby/object:Settings bowtieIndex: /usr/share/hgChrAll humanGenomeDatabase: /usr/share novoIndex: /usr/share/hgChrAll.ndx ntDatabase: /usr/share/nt/nt
In addition, configuation files are produced for each sequence in
$(pwd)/.sequenceName. Settings for each program are chosen by default but can be changed via the sequence file which has the following form:
--- !ruby/object:Sequence abyssPath: s_4_sequence_Akata.txt.novo.NM.fasta.nospans.blast.megan.rma abyssPathGlob: reads-*.fasta blast1Path: s_4_sequence_Akata.txt.novo.NM.fasta.nospans blast2Path: s_4_sequence_Akata.txt.novo.NM.fasta.nospans.blast.megan.rma.kmerOptimized.fa blastOutputFormat: 0 blastPathGlob: reads-*.fasta.*.kmer.contigs.fa.kmerOptimized.fa dataType: illumina1.3 eValue1: 1.0e-06 eValue2: 100 expansionNumber: 10 filePath: s_4_sequence_Akata.txt imageFileType: jpg maxKmerLength: 38 maxMatches: 0 megan1Path: s_4_sequence_Akata.txt.novo.NM.fasta.nospans.blast megan2Path: s_4_sequence_Akata.txt.novo.NM.fasta.nospans.blast.megan.rma.kmerOptimized.fa.blast minKmerLength: 7 minScoreByLength: 0 minSupport: 5 novoalignPath: s_4_sequence_Akata.txt.novo pipeEndPath: s_4_sequence_Akata.txt.novo.NM.fasta.nospans.blast.megan.rma.kmerOptimized.fa.blast.megan.rma readLength: 38 removeNonMappedPath: s_4_sequence_Akata.txt.novo.NM.fasta topPercent: 10.0 useCogs: "false" useGos: "false" winScore: 0.0
It is recommended the path variables not be adjusted. Everything except expansionNumber, which is the number of times to execute the expansion command when generating a picture from MEGAN is straight-forward.
In addition, the amount of RAM, number of CPU cores, CPU architecture, operating system, default shell, and existence of locate database is automatically computed each execution but not stored.