How to execute FusionSeq - Revision history

Asboner: /* Junction-Sequence Identifier module */

2011-04-05T17:59:52Z

Junction-Sequence Identifier module

← Older revision		Revision as of 17:59, 5 April 2011
Line 61:		Line 61:
	==Junction-Sequence Identifier module==		==Junction-Sequence Identifier module==
	The junction-sequence identifier uses the high-confidence gfr file and look for the sequence of the junction for each candidate. This is the most computationally expensive part of FusionSeq. In order to run it efficiently we exploit a parallel computing architecture. Typically, one would execute:		The junction-sequence identifier uses the high-confidence gfr file and look for the sequence of the junction for each candidate. This is the most computationally expensive part of FusionSeq. In order to run it efficiently we exploit a parallel computing architecture. Typically, one would execute:
	$ gfr2bpJunctions data.confidence.gfr 40 200		$ gfr2bpJunctions data.confidence.gfr 40 200 1.0
	where 40 correspond to the tile size and 200 is the number of nucleotides flanking the exons. For example, with 50-mers, having 40 as tile size ensures that at least 10 nucleotides of the reads will be mapped to either of the tiles. This program generates the fusion junction library in fasta format for each candidate, split in several files each one containing at most 2M entries. It also creates two additional files: one (data_jobList1.txt) including the instructions to index the library and align the reads to the files, and the second (data_jobList2.txt) including the instructions to aggregate the results of the alignment. How to run those jobs depends on the user architecture of the cluster. It is worth noting that data_jobList1.txt expects a compressed file with all the single end reads called: data_allReads.txt.gz. The auxiliary program [[FusionSeq_List of programs#export2mrf\|export2mrf]] generates this file automatically from the output files (export) of the Illumina platform. Otherwise, the user should provide a file with all the reads, one sequence per line. One way to do this from the fastq data is:		where 40 correspond to the tile size and 200 is the number of nucleotides flanking the exons. For example, with 50-mers, having 40 as tile size ensures that at least 10 nucleotides of the reads will be mapped to either of the tiles. This program generates the fusion junction library in fasta format for each candidate with DASPER>1.0, split in several files each one containing at most 2M entries. It also creates two additional files: one (data_jobList1.txt) including the instructions to index the library and align the reads to the files, and the second (data_jobList2.txt) including the instructions to aggregate the results of the alignment. How to run those jobs depends on the user architecture of the cluster. It is worth noting that data_jobList1.txt expects a compressed file with all the single end reads called: data_allReads.txt.gz. The auxiliary program [[FusionSeq_List of programs#export2mrf\|export2mrf]] generates this file automatically from the output files (export) of the Illumina platform. Otherwise, the user should provide a file with all the reads, one sequence per line. One way to do this from the fastq data is:
	cat data_1.fastq data_2.fastq \| gawk '$1 ~/^@/ { getline; print }' - > data_allReads.txt		cat data_1.fastq data_2.fastq \| gawk '$1 ~/^@/ { getline; print }' - > data_allReads.txt
	where data_1.fastq data_2.fastq are the fastq data of the two ends.		where data_1.fastq data_2.fastq are the fastq data of the two ends.

Asboner at 11:14, 17 February 2011

2011-02-17T11:14:30Z

← Older revision		Revision as of 11:14, 17 February 2011
Line 4:		Line 4:
	For data generated by the Genome Analyzer II and aligned with Eland, the program [[FusionSeq_List_of_programs#export2mrf\|export2mrf]] provided with FusionSeq can be used to perform the conversion. All programs provide a brief explanation of their usage when run without parameters. A typical analysis session may look like the following:		For data generated by the Genome Analyzer II and aligned with Eland, the program [[FusionSeq_List_of_programs#export2mrf\|export2mrf]] provided with FusionSeq can be used to perform the conversion. All programs provide a brief explanation of their usage when run without parameters. A typical analysis session may look like the following:
	<pre>		<pre>
	$ geneFusions data 4 < data.mrf > data.1.gfr 2> data.1.gfr.log		$ geneFusions data 4 < data.mrf \| gfrClassify > data.1.gfr 2> data.1.gfr.log

	$ (gfrMitochondrialFilter < data.1.gfr \| gfrRepeatMaskerFilter repeatMasker.interval 5 \| gfrCountPairTypes \| gfrExpressionConsistencyFilter \|		$ (gfrMitochondrialFilter < data.1.gfr \| gfrRepeatMaskerFilter repeatMasker.interval 5 \| gfrCountPairTypes \| gfrExpressionConsistencyFilter \|

Asboner: /* Auxilliary data */

2011-02-16T11:18:05Z

Auxilliary data

← Older revision		Revision as of 11:18, 16 February 2011
Line 55:		Line 55:
	If the auxilliary modules are installed, then, the fourth command generates the corresponding files and print the gfr file to standard output.		If the auxilliary modules are installed, then, the fourth command generates the corresponding files and print the gfr file to standard output.
	$ (gfr2images < data.confidence.gfr \| gfr2bed \| gfr2fasta \| gfr2gff) 2> data.aux.log		$ (gfr2images < data.confidence.gfr \| gfr2bed \| gfr2fasta \| gfr2gff) 2> data.aux.log
	The files needs to be properly located into the directory structure described in Installing CGIs.		The files needs to be properly located into the directory structure described in Installing CGIs. Also, note that [[#gfrCountPairTypes\|gfrCountPairTypes]] should be executed before [[#gfr2images\|gfr2images]].

	<center>[[#top\|Top]]</center>		<center>[[#top\|Top]]</center>

Asboner at 06:44, 3 December 2010

2010-12-03T06:44:11Z

← Older revision		Revision as of 06:44, 3 December 2010
Line 6:		Line 6:
	$ geneFusions data 4 < data.mrf > data.1.gfr 2> data.1.gfr.log		$ geneFusions data 4 < data.mrf > data.1.gfr 2> data.1.gfr.log

	$ (gfrMitochondrialFilter < data.1.gfr \| gfrRepeatMaskerFilter ~~/path/to/~~repeatMasker.interval 5 \| gfrCountPairTypes \| gfrExpressionConsistencyFilter \|		$ (gfrMitochondrialFilter < data.1.gfr \| gfrRepeatMaskerFilter repeatMasker.interval 5 \| gfrCountPairTypes \| gfrExpressionConsistencyFilter \|
	gfrAbnormalInsertSizeFilter 0.01 \| gfrPCRFilter 4 4 \| gfrProximityFilter 1000 \| gfrAddInfo \| gfrAnnotationConsistencyFilter ribosomal \|		gfrAbnormalInsertSizeFilter 0.01 \| gfrPCRFilter 4 4 \| gfrProximityFilter 1000 \| gfrAddInfo \| gfrAnnotationConsistencyFilter ribosomal \|
	gfrAnnotationConsistencyFilter pseudogenes \| gfrBlackListFilter blackList.txt \| gfrLargeScaleHomologyFilter \|		gfrAnnotationConsistencyFilter pseudogenes \| gfrBlackListFilter blackList.txt \| gfrLargeScaleHomologyFilter \|
Line 36:		Line 36:
	The second command:		The second command:
	<pre>		<pre>
	$ (gfrMitochondrialFilter < data.1.gfr \| gfrRepeatMaskerFilter ~~/path/to/~~repeatMasker.interval 5 \| gfrCountPairTypes \| gfrExpressionConsistencyFilter \|		$ (gfrMitochondrialFilter < data.1.gfr \| gfrRepeatMaskerFilter repeatMasker.interval 5 \| gfrCountPairTypes \| gfrExpressionConsistencyFilter \|
	gfrAbnormalInsertSizeFilter 0.01 \| gfrPCRFilter 4 4 \| gfrProximityFilter 1000 \| gfrAddInfo \| gfrAnnotationConsistencyFilter ribosomal \|		gfrAbnormalInsertSizeFilter 0.01 \| gfrPCRFilter 4 4 \| gfrProximityFilter 1000 \| gfrAddInfo \| gfrAnnotationConsistencyFilter ribosomal \|
	gfrAnnotationConsistencyFilter pseudogenes \| gfrBlackListFilter blackList.txt \| gfrLargeScaleHomologyFilter \|		gfrAnnotationConsistencyFilter pseudogenes \| gfrBlackListFilter blackList.txt \| gfrLargeScaleHomologyFilter \|

Asboner: /* Filtration cascade module */

2010-12-03T06:37:18Z

Filtration cascade module

← Older revision		Revision as of 06:37, 3 December 2010
Line 34:		Line 34:

	==Filtration cascade module==		==Filtration cascade module==
	The second command		The second command:
	<pre>		<pre>
	$ (~~gfrAbnormalInsertSizeFilter 0.01~~ < data.1.gfr \| gfrPCRFilter 4 4 \| gfrProximityFilter 1000 \| gfrAddInfo \|		$ (gfrMitochondrialFilter < data.1.gfr \| gfrRepeatMaskerFilter /path/to/repeatMasker.interval 5 \| gfrCountPairTypes \| gfrExpressionConsistencyFilter \|
	gfrAnnotationConsistencyFilter ~~ribosomal~~ \| gfrBlackListFilter blackList.txt \| gfrLargeScaleHomologyFilter \|		gfrAbnormalInsertSizeFilter 0.01 \| gfrPCRFilter 4 4 \| gfrProximityFilter 1000 \| gfrAddInfo \| gfrAnnotationConsistencyFilter ribosomal \|
	gfrRibosomalFilter \| gfrSmallScaleHomologyFilter) > data.gfr 2> data.log		gfrAnnotationConsistencyFilter pseudogenes \| gfrBlackListFilter blackList.txt \| gfrLargeScaleHomologyFilter \|
			gfrRibosomalFilter \| gfrSmallScaleHomologyFilter) > data.gfr 2> data.log
	</pre>		</pre>
	runs several filters to remove potential artifacts and generate a high-confidence list of fusion candidates. The order of the filters affects only the computation time of the processing. However, [[FusionSeq_List of programs#gfrAddInfo\|gfrAddInfo]] needs to be executed prior to [[FusionSeq_List of programs#gfrAnnotationConsistencyFilter\|gfrAnnotationConsistencyFilter]] or [[FusionSeq_List of programs#gfrBlacklistFilter\|gfrBlackListFilter]], because they require gene symbols and descriptions.		runs several filters to remove potential artifacts and generate a high-confidence list of fusion candidates. The order of the filters affects only the computation time of the processing. However, [[FusionSeq_List of programs#gfrAddInfo\|gfrAddInfo]] needs to be executed prior to [[FusionSeq_List of programs#gfrAnnotationConsistencyFilter\|gfrAnnotationConsistencyFilter]] or [[FusionSeq_List of programs#gfrBlacklistFilter\|gfrBlackListFilter]], because they require gene symbols and descriptions.

Asboner at 06:36, 3 December 2010

2010-12-03T06:36:44Z

← Older revision		Revision as of 06:36, 3 December 2010
Line 6:		Line 6:
	$ geneFusions data 4 < data.mrf > data.1.gfr 2> data.1.gfr.log		$ geneFusions data 4 < data.mrf > data.1.gfr 2> data.1.gfr.log

	$ (~~gfrAbnormalInsertSizeFilter 0.01~~ < data.1.gfr \| gfrPCRFilter 4 4 \| gfrProximityFilter 1000 \| gfrAddInfo \|		$ (gfrMitochondrialFilter < data.1.gfr \| gfrRepeatMaskerFilter /path/to/repeatMasker.interval 5 \| gfrCountPairTypes \| gfrExpressionConsistencyFilter \|
	gfrAnnotationConsistencyFilter ~~ribosomal~~ \| gfrBlackListFilter blackList.txt \| gfrLargeScaleHomologyFilter \|		gfrAbnormalInsertSizeFilter 0.01 \| gfrPCRFilter 4 4 \| gfrProximityFilter 1000 \| gfrAddInfo \| gfrAnnotationConsistencyFilter ribosomal \|
	gfrRibosomalFilter \| gfrSmallScaleHomologyFilter) > data.gfr 2> data.log		gfrAnnotationConsistencyFilter pseudogenes \| gfrBlackListFilter blackList.txt \| gfrLargeScaleHomologyFilter \|
			gfrRibosomalFilter \| gfrSmallScaleHomologyFilter) > data.gfr 2> data.log

	$ gfrConfidenceValues data < data.gfr > data.confidence.gfr		$ gfrConfidenceValues data < data.gfr > data.confidence.gfr

Asboner at 06:51, 21 November 2010

2010-11-21T06:51:49Z

← Older revision		Revision as of 06:51, 21 November 2010
Line 18:		Line 18:
	$ validateBpJunctions < data_00002_AB.bp \| bpFilter 4 4 100 0.01 30 > data_00002_AB.filtered.bp		$ validateBpJunctions < data_00002_AB.bp \| bpFilter 4 4 100 0.01 30 > data_00002_AB.filtered.bp

	$ bp2alignment data_00002_AB.filtered.bp > data_00002_breakPointAlignments.txt		$ bp2alignment < data_00002_AB.filtered.bp > data_00002_breakPointAlignments.txt

	$ bp2wig data_00002_AB.filtered.bp		$ bp2wig data_00002_AB.filtered.bp

Asboner at 06:50, 21 November 2010

2010-11-21T06:50:29Z

← Older revision		Revision as of 06:50, 21 November 2010
Line 4:		Line 4:
	For data generated by the Genome Analyzer II and aligned with Eland, the program [[FusionSeq_List_of_programs#export2mrf\|export2mrf]] provided with FusionSeq can be used to perform the conversion. All programs provide a brief explanation of their usage when run without parameters. A typical analysis session may look like the following:		For data generated by the Genome Analyzer II and aligned with Eland, the program [[FusionSeq_List_of_programs#export2mrf\|export2mrf]] provided with FusionSeq can be used to perform the conversion. All programs provide a brief explanation of their usage when run without parameters. A typical analysis session may look like the following:
	<pre>		<pre>
	$ geneFusions data 4 < data.mrf > data.1.gfr 2> data.1.log		$ geneFusions data 4 < data.mrf > data.1.gfr 2> data.1.gfr.log

	$ (gfrAbnormalInsertSizeFilter 0.01 < data.1.gfr \| gfrPCRFilter 4 4 \| gfrProximityFilter 1000 \| gfrAddInfo \|		$ (gfrAbnormalInsertSizeFilter 0.01 < data.1.gfr \| gfrPCRFilter 4 4 \| gfrProximityFilter 1000 \| gfrAddInfo \|

Asboner: /* Fusion detection module */

2010-11-21T06:50:02Z

Fusion detection module

← Older revision		Revision as of 06:50, 21 November 2010
Line 24:		Line 24:

	==Fusion detection module==		==Fusion detection module==
	The first command ([[FusionSeq_List of programs#geneFusions\|geneFusions]]) will create the first list of candidate fusion transcripts. The ~~parameters~~ "data" ~~and "4" correspond~~ to the prefix ~~(data)~~ that will be used to generate the IDs of the candidates, namely data_00001, data_00002, etc. "4" is the minimum number of PE reads needed to call a candidate. The program reads data.mrf from standard input and save the output to data.1.gfr as well as logging information on data.1.log. Note that all programs will generate some logging information in this format:		The first command ([[FusionSeq_List of programs#geneFusions\|geneFusions]]) will create the first list of candidate fusion transcripts:
			$ geneFusions data 4 < data.mrf > data.1.gfr 2> data.1.gfr.log
			The parameter "data" corresponds to the prefix that will be used to generate the IDs of the candidates, namely data_00001, data_00002, etc., whereas "4" is the minimum number of inter-transcript PE reads needed to call a candidate. The program reads data.mrf from standard input and save the output to data.1.gfr as well as logging information on data.1.log. Note that all programs will generate some logging information in this format:
	WARNING: <program.name>_<type.of.information>: <value>		WARNING: <program.name>_<type.of.information>: <value>
	Hence we recommend to always capture the LOG information in a file by redirecting the STDERR with '2>'.		Hence we recommend to always capture the LOG information in a file by redirecting the STDERR with '2>'.

	The ~~output in~~ data.1.gfr ~~can be considered~~ the most comprehensive list of fusion candidates.		The content of data.1.gfr is the most comprehensive list of fusion candidates.

	==Filtration cascade module==		==Filtration cascade module==
	The second command		The second command

Asboner: /* Junction-Sequence Identifier module */

2010-11-20T18:37:10Z

Junction-Sequence Identifier module

← Older revision		Revision as of 18:37, 20 November 2010
Line 65:		Line 65:
	$ validateBpJunctions < data_00002_AB.bp \| bpFilter 4 4 100 0.01 30 > data_00002_AB.filtered.bp		$ validateBpJunctions < data_00002_AB.bp \| bpFilter 4 4 100 0.01 30 > data_00002_AB.filtered.bp

	$ bp2alignment data_00002_AB.filtered.bp > data_00002_breakPointAlignments.txt		$ bp2alignment < data_00002_AB.filtered.bp > data_00002_breakPointAlignments.txt

	$ bp2wig data_00002_AB.filtered.bp		$ bp2wig data_00002_AB.filtered.bp