BRAKER-Sapelo2 - Revision history

Shtsai at 00:17, 5 April 2026

2026-04-05T00:17:20Z

← Older revision		Revision as of 20:17, 4 April 2026
Line 1:		Line 1:
	[[Category:Sapelo2]][[Category:Software]][[Category:Bioinformatics]]		[[Category:Sapelo2]][[Category:Software]][[Category:Bioinformatics]]
	=== Category ===		== Category ==

	Bioinformatics		Bioinformatics

	=== Program On ===		== Program On ==

	Sapelo2		Sapelo2

	=== Version ===		== Version ==
	3.0.8		3.0.8

	=== Author / Distributor ===		== Author / Distributor ==
	[https://github.com/Gaius-Augustus/BRAKER Gaius-Augustus]		[https://github.com/Gaius-Augustus/BRAKER Gaius-Augustus]

	=== Description ===		== Description ==
	"BRAKER is a pipeline for fully automated prediction of protein coding genes with GeneMark-ES/ET and AUGUSTUS in novel eukaryotic genomes."		"BRAKER is a pipeline for fully automated prediction of protein coding genes with GeneMark-ES/ET and AUGUSTUS in novel eukaryotic genomes."
	More details are at [https://github.com/Gaius-Augustus/BRAKER BRAKER]		More details are at [https://github.com/Gaius-Augustus/BRAKER BRAKER]

	=== Running Program ===		== Running Program ==

	Also refer to [[Running Jobs on Sapelo2]]		Also refer to [[Running Jobs on Sapelo2]]
Line 106:		Line 106:
	</pre>		</pre>

	=== Documentation ===		== Documentation ==

	<pre class="gcommand">		<pre class="gcommand">
Line 534:		Line 534:
	[[#top\|Back to Top]]		[[#top\|Back to Top]]

	=== Installation ===		== Installation ==

	source code from [https://github.com/Gaius-Augustus/BRAKER BRAKER]		source code from [https://github.com/Gaius-Augustus/BRAKER BRAKER]

	=== System ===		== System ==
	64-bit Linux		64-bit Linux

Chelsea at 14:36, 8 July 2025

2025-07-08T14:36:20Z

← Older revision		Revision as of 10:36, 8 July 2025
Line 9:		Line 9:

	=== Version ===		=== Version ===
	~~2.1.6, 3.0.3,~~ 3.0.8		3.0.8

	=== Author / Distributor ===		=== Author / Distributor ===
Line 22:		Line 22:
	Also refer to [[Running Jobs on Sapelo2]]		Also refer to [[Running Jobs on Sapelo2]]

			<!--
	'''Version 2.1.6'''		'''Version 2.1.6'''

Line 58:		Line 58:
	module load BRAKER/3.0.3-foss-2022a-long_reads		module load BRAKER/3.0.3-foss-2022a-long_reads
	</div>		</div>
			-->

	'''Version 3.0.8'''		'''Version 3.0.8'''
Line 63:		Line 64:
	Version 3.0.8 installed at		Version 3.0.8 installed at

	* /apps/eb/BRAKER/3.0.8-foss-~~2022a~~/		* /apps/eb/BRAKER/3.0.8-foss-2023a/

	To use it, please load the module with:		To use it, please load the module with:
	<div class="gscript2">		<div class="gscript2">
	module load BRAKER/3.0.8-foss-~~2022a~~		module load BRAKER/3.0.8-foss-2023a
	</div>		</div>

Line 108:		Line 109:

	<pre class="gcommand">		<pre class="gcommand">
	module load BRAKER/3.0.8-foss-~~2022a~~		module load BRAKER/3.0.8-foss-2023a
	braker.pl		braker.pl

Jordan: /* Running Program */

2025-05-06T17:13:57Z

Running Program

← Older revision		Revision as of 13:13, 6 May 2025
Line 73:		Line 73:
	'''Notes about Dependencies, Etc.'''		'''Notes about Dependencies, Etc.'''

	BRAKER is a gene prediction pipeline that predicts the genes in an input genome using either just the genome itself (referred to as "<i>ab initio</i> mode but in more modern parlance, the equivalent of <i>de novo</i>), or the input genome with RNAseq data, protein data, or a combination of RNAseq and protein data to support and solidify gene predictions. If it detects the input of both RNA and proteins as supporting data, BRAKER will implement GeneMark-ETP rather than the other iterations of GeneMark; one of the scripts that GeneMark-ETP uses to filter results calls DIAMOND (unless BLAST is explicitly specified by the user) without passing the value of the user-specified "--threads" option (see full documentation below) to it, which will cause DIAMOND to auto-detect the number of CPUs present on the node and attempt to run this many threads regardless of how many CPUs have been allocated for the job. Therefore, it is <i>strongly recommended</i> that if you run BRAKER with both RNA and protein data, you request a full 128 CPUs for your BRAKER job; while this may cause your job to wait in the queue for longer, numerous BRAKER jobs have been observed to get "stuck" or outright fail because of DIAMOND attempting to run 128 threads on only a few CPUs. <b>Regardless of the type(s) of input data you use, BRAKER will use one iteration or another of the GeneMark gene prediction software, and thus in order to use BRAKER, you need to have downloaded the key file for GeneMarker and put it into your home directory. Please follow instructions found at <u>[[GeneMarkES-Sapelo2]]</u>.</b>		BRAKER is a gene prediction pipeline that predicts the genes in an input genome using either just the genome itself (referred to as "<i>ab initio</i>" mode but in more modern parlance, the equivalent of <i>de novo</i>), or the input genome with RNAseq data, protein data, or a combination of RNAseq and protein data to support and solidify gene predictions. If it detects the input of both RNA and proteins as supporting data, BRAKER will implement GeneMark-ETP rather than the other iterations of GeneMark; one of the scripts that GeneMark-ETP uses to filter results calls DIAMOND (unless BLAST is explicitly specified by the user) without passing the value of the user-specified "--threads" option (see full documentation below) to it, which will cause DIAMOND to auto-detect the number of CPUs present on the node and attempt to run this many threads regardless of how many CPUs have been allocated for the job. Therefore, it is <i>strongly recommended</i> that if you run BRAKER with both RNA and protein data, you request a full 128 CPUs for your BRAKER job; while this may cause your job to wait in the queue for longer, numerous BRAKER jobs have been observed to get "stuck" or outright fail because of DIAMOND attempting to run 128 threads on only a few CPUs. <b>Regardless of the type(s) of input data you use, BRAKER will use one iteration or another of the GeneMark gene prediction software, and thus in order to use BRAKER, you need to have downloaded the key file for GeneMarker and put it into your home directory. Please follow instructions found at <u>[[GeneMarkES-Sapelo2]]</u>.</b>

	Furthermore, BRAKER makes use of numerous external tools (as can be seen in the "CONFIGURATION OPTIONS" section of the documentation below); while working versions of these are all included with the modules listed above, you also have the option to use your own versions of these (see the example submission script below):		Furthermore, BRAKER makes use of numerous external tools (as can be seen in the "CONFIGURATION OPTIONS" section of the documentation below); while working versions of these are all included with the modules listed above, you also have the option to use your own versions of these (see the example submission script below):

Jordan: /* Running Program */

2025-05-06T17:13:37Z

Running Program

← Older revision		Revision as of 13:13, 6 May 2025
Line 73:		Line 73:
	'''Notes about Dependencies, Etc.'''		'''Notes about Dependencies, Etc.'''

	BRAKER is a gene prediction pipeline that predicts the genes in an input genome using RNAseq data, protein data, or a combination of RNAseq and protein data to support and solidify gene predictions. If it detects the input of both ~~kinds of~~ supporting data, BRAKER will implement GeneMark-ETP rather than the other iterations of GeneMark; one of the scripts that GeneMark-ETP uses to filter results calls DIAMOND (unless BLAST is explicitly specified by the user) without passing the value of the user-specified "--threads" option (see full documentation below) to it, which will cause DIAMOND to auto-detect the number of CPUs present on the node and attempt to run this many threads regardless of how many CPUs have been allocated for the job. Therefore, it is <i>strongly recommended</i> that if you run BRAKER with both RNA and protein data, you request a full 128 CPUs for your BRAKER job; while this may cause your job to wait in the queue for longer, numerous BRAKER jobs have been observed to get "stuck" or outright fail because of DIAMOND attempting to run 128 threads on only a few CPUs. <b>Regardless of the type(s) of input data you use, BRAKER will use one iteration or another of the GeneMark gene prediction software, and thus in order to use BRAKER, you need to have downloaded the key file for GeneMarker and put it into your home directory. Please follow instructions found at <u>[[GeneMarkES-Sapelo2]]</u>.</b>		BRAKER is a gene prediction pipeline that predicts the genes in an input genome using either just the genome itself (referred to as "<i>ab initio</i> mode but in more modern parlance, the equivalent of <i>de novo</i>), or the input genome with RNAseq data, protein data, or a combination of RNAseq and protein data to support and solidify gene predictions. If it detects the input of both RNA and proteins as supporting data, BRAKER will implement GeneMark-ETP rather than the other iterations of GeneMark; one of the scripts that GeneMark-ETP uses to filter results calls DIAMOND (unless BLAST is explicitly specified by the user) without passing the value of the user-specified "--threads" option (see full documentation below) to it, which will cause DIAMOND to auto-detect the number of CPUs present on the node and attempt to run this many threads regardless of how many CPUs have been allocated for the job. Therefore, it is <i>strongly recommended</i> that if you run BRAKER with both RNA and protein data, you request a full 128 CPUs for your BRAKER job; while this may cause your job to wait in the queue for longer, numerous BRAKER jobs have been observed to get "stuck" or outright fail because of DIAMOND attempting to run 128 threads on only a few CPUs. <b>Regardless of the type(s) of input data you use, BRAKER will use one iteration or another of the GeneMark gene prediction software, and thus in order to use BRAKER, you need to have downloaded the key file for GeneMarker and put it into your home directory. Please follow instructions found at <u>[[GeneMarkES-Sapelo2]]</u>.</b>

	Furthermore, BRAKER makes use of numerous external tools (as can be seen in the "CONFIGURATION OPTIONS" section of the documentation below); while working versions of these are all included with the modules listed above, you also have the option to use your own versions of these (see the example submission script below):		Furthermore, BRAKER makes use of numerous external tools (as can be seen in the "CONFIGURATION OPTIONS" section of the documentation below); while working versions of these are all included with the modules listed above, you also have the option to use your own versions of these (see the example submission script below):

Jordan: /* Running Program */

2025-05-06T17:06:55Z

Running Program

← Older revision		Revision as of 13:06, 6 May 2025
Line 73:		Line 73:
	'''Notes about Dependencies, Etc.'''		'''Notes about Dependencies, Etc.'''

	BRAKER is a gene prediction pipeline that predicts the genes in an input genome using RNAseq data, protein data, or a combination of RNAseq and protein data to support and solidify gene predictions. If it detects the input of both kinds of supporting data, BRAKER will implement GeneMark-ETP rather than the other iterations of GeneMark; one of the scripts that GeneMark-ETP uses to filter results calls DIAMOND (unless BLAST is explicitly specified by the user) without passing the value of the user-specified "--threads" option (see full documentation below) to it, which will cause DIAMOND to auto-detect the number of CPUs present on the node and attempt to run this many threads regardless of how many CPUs have been allocated for the job. Therefore, it is <i>strongly recommended</i> that if you run BRAKER with both RNA and protein data, you request a full 128 CPUs for your BRAKER job; while this may cause your job to wait in the queue for longer, numerous BRAKER jobs have been observed to get "stuck" or outright fail because of DIAMOND attempting to run 128 threads on only a few CPUs. <b>Regardless of the type(s) of input data you use, BRAKER will use one iteration or another of the GeneMark gene prediction software, and thus in order to use BRAKER, you need to have downloaded the key file for GeneMarker and put it into your home directory. Please follow instructions found at [[GeneMarkES-Sapelo2]].</b>		BRAKER is a gene prediction pipeline that predicts the genes in an input genome using RNAseq data, protein data, or a combination of RNAseq and protein data to support and solidify gene predictions. If it detects the input of both kinds of supporting data, BRAKER will implement GeneMark-ETP rather than the other iterations of GeneMark; one of the scripts that GeneMark-ETP uses to filter results calls DIAMOND (unless BLAST is explicitly specified by the user) without passing the value of the user-specified "--threads" option (see full documentation below) to it, which will cause DIAMOND to auto-detect the number of CPUs present on the node and attempt to run this many threads regardless of how many CPUs have been allocated for the job. Therefore, it is <i>strongly recommended</i> that if you run BRAKER with both RNA and protein data, you request a full 128 CPUs for your BRAKER job; while this may cause your job to wait in the queue for longer, numerous BRAKER jobs have been observed to get "stuck" or outright fail because of DIAMOND attempting to run 128 threads on only a few CPUs. <b>Regardless of the type(s) of input data you use, BRAKER will use one iteration or another of the GeneMark gene prediction software, and thus in order to use BRAKER, you need to have downloaded the key file for GeneMarker and put it into your home directory. Please follow instructions found at <u>[[GeneMarkES-Sapelo2]]</u>.</b>

	Furthermore, BRAKER makes use of numerous external tools (as can be seen in the "CONFIGURATION OPTIONS" section of the documentation below); while working versions of these are all included with the modules listed above, you also have the option to use your own versions of these (see the example submission script below):		Furthermore, BRAKER makes use of numerous external tools (as can be seen in the "CONFIGURATION OPTIONS" section of the documentation below); while working versions of these are all included with the modules listed above, you also have the option to use your own versions of these (see the example submission script below):

Jordan: /* Running Program */

2025-04-24T20:07:22Z

Running Program

← Older revision		Revision as of 16:07, 24 April 2025
Line 70:		Line 70:
	</div>		</div>

	In order to use BRAKER, you need to have downloaded the key file for GeneMarker and put it into your home directory. Please follow instructions found at [[GeneMarkES-Sapelo2]].
			'''Notes about Dependencies, Etc.'''

			BRAKER is a gene prediction pipeline that predicts the genes in an input genome using RNAseq data, protein data, or a combination of RNAseq and protein data to support and solidify gene predictions. If it detects the input of both kinds of supporting data, BRAKER will implement GeneMark-ETP rather than the other iterations of GeneMark; one of the scripts that GeneMark-ETP uses to filter results calls DIAMOND (unless BLAST is explicitly specified by the user) without passing the value of the user-specified "--threads" option (see full documentation below) to it, which will cause DIAMOND to auto-detect the number of CPUs present on the node and attempt to run this many threads regardless of how many CPUs have been allocated for the job. Therefore, it is <i>strongly recommended</i> that if you run BRAKER with both RNA and protein data, you request a full 128 CPUs for your BRAKER job; while this may cause your job to wait in the queue for longer, numerous BRAKER jobs have been observed to get "stuck" or outright fail because of DIAMOND attempting to run 128 threads on only a few CPUs. <b>Regardless of the type(s) of input data you use, BRAKER will use one iteration or another of the GeneMark gene prediction software, and thus in order to use BRAKER, you need to have downloaded the key file for GeneMarker and put it into your home directory. Please follow instructions found at [[GeneMarkES-Sapelo2]].</b>

	Furthermore, BRAKER makes use of numerous external tools (as can be seen in the "CONFIGURATION OPTIONS" section of the documentation below); while working versions of these are all included with the modules listed above, you also have the option to use your own versions of these (see the example submission script below):		Furthermore, BRAKER makes use of numerous external tools (as can be seen in the "CONFIGURATION OPTIONS" section of the documentation below); while working versions of these are all included with the modules listed above, you also have the option to use your own versions of these (see the example submission script below):

Jordan at 20:54, 14 June 2024

2024-06-14T20:54:19Z

← Older revision		Revision as of 16:54, 14 June 2024
Line 91:		Line 91:

	module load BRAKER/3.0.8-foss-2022a<br>		module load BRAKER/3.0.8-foss-2022a<br>
	~~module load TSEBRA/1.1.2-foss-2022a~~

	braker.pl --genome=Newly_assembled_genome.fasta --prot_seq=Protein_database.faa --PROTHINT_PATH=<u>/Path/to/MY/version/of/ProtHint</u><br>		braker.pl --genome=Newly_assembled_genome.fasta --prot_seq=Protein_database.faa --PROTHINT_PATH=<u>/Path/to/MY/version/of/ProtHint</u><br>

Jordan: Updated to reflect that all necessary tools (including ProtHint and TSEBRA) are now included in BRAKER/3.0.8-foss-2022a

2024-06-14T18:22:09Z

Updated to reflect that all necessary tools (including ProtHint and TSEBRA) are now included in BRAKER/3.0.8-foss-2022a

← Older revision		Revision as of 14:22, 14 June 2024
Line 72:		Line 72:
	In order to use BRAKER, you need to have downloaded the key file for GeneMarker and put it into your home directory. Please follow instructions found at [[GeneMarkES-Sapelo2]].		In order to use BRAKER, you need to have downloaded the key file for GeneMarker and put it into your home directory. Please follow instructions found at [[GeneMarkES-Sapelo2]].

	Furthermore, BRAKER ~~requires TSEBRA,~~ of ~~which Version 1.1.2 is installed on Sapelo2 but not currently~~ included ~~in BRAKER/3.0.8-foss-2022a. Therefore~~, you ~~will~~ also ~~need to load TSEBRA separately if you want~~ to use ~~Version 3.0.8~~ of ~~BRAKER. TSEBRA 1.1.2 can be found at~~		Furthermore, BRAKER makes use of numerous external tools (as can be seen in the "CONFIGURATION OPTIONS" section of the documentation below); while working versions of these are all included with the modules listed above, you also have the option to use your own versions of these (see the example submission script below):

	* /apps/eb/TSEBRA/1.1.2-foss-2022a-long_reads/		Here is an example of a shell script sub.sh to run BRAKER using a protein database from the batch queue along with your own version of ProtHint (in place of the default one, which does not need to be specified):
	* /apps/eb/TSEBRA/1.1.2-foss-2022a/
	* /apps/eb/TSEBRA/1.1.2-GCCcore-11.3.0

	~~It is recommended that you use the second option above, so as to avoid any dependency conflicts or other unforeseeable problems.~~

	Finally, in order to use BRAKER to predict genes using a protein database, BRAKER needs to use ProtHint; this is included in GeneMark, but BRAKER does not always successfully find it, so you may find it useful to specify the path to it using the following (see also the example submission script below):
	~~<pre class="gcommand">~~
	~~--PROTHINT_PATH=/apps/eb/GeneMark-ET/4.71-GCCcore-11.3.0/ProtHint/bin/~~
	~~</pre>~~

	Here is an example of a shell script sub.sh to run BRAKER using a protein database from the batch queue:

	<div class="gscript2">		<div class="gscript2">
Line 104:		Line 93:
	module load TSEBRA/1.1.2-foss-2022a		module load TSEBRA/1.1.2-foss-2022a

	braker.pl --genome=Newly_assembled_genome.fasta --prot_seq=Protein_database.faa --PROTHINT_PATH=/~~apps~~/eb/~~GeneMark-ET~~/~~4.71-GCCcore-11.3.0~~/ProtHint/~~bin/~~		braker.pl --genome=Newly_assembled_genome.fasta --prot_seq=Protein_database.faa --PROTHINT_PATH=<u>/Path/to/MY/version/of/ProtHint</u><br>
	</div>		</div>

	In the real submission script, be sure to replace the above underlined values with proper values.		In the real submission script, be sure to replace the above <u>underlined</u> values with proper values.

	Here is an example of job submission command:		Here is an example of job submission command:

Jordan: Revised the recommendations regarding TSEBRA, it turns out that this actually does get loaded with the two foss-2022a modules for v3.0.3

2024-06-14T14:39:19Z

Revised the recommendations regarding TSEBRA, it turns out that this actually does get loaded with the two foss-2022a modules for v3.0.3

← Older revision		Revision as of 10:39, 14 June 2024
Line 72:		Line 72:
	In order to use BRAKER, you need to have downloaded the key file for GeneMarker and put it into your home directory. Please follow instructions found at [[GeneMarkES-Sapelo2]].		In order to use BRAKER, you need to have downloaded the key file for GeneMarker and put it into your home directory. Please follow instructions found at [[GeneMarkES-Sapelo2]].

	Furthermore, BRAKER requires TSEBRA, of which Version 1.1.2 is installed on Sapelo2 but not currently included in ~~any of the~~ BRAKER ~~modules we have installed~~. Therefore, you will also need to load TSEBRA ~~when~~ you ~~load~~ BRAKER. TSEBRA 1.1.2 can be found at		Furthermore, BRAKER requires TSEBRA, of which Version 1.1.2 is installed on Sapelo2 but not currently included in BRAKER/3.0.8-foss-2022a. Therefore, you will also need to load TSEBRA separately if you want to use Version 3.0.8 of BRAKER. TSEBRA 1.1.2 can be found at

	* /apps/eb/TSEBRA/1.1.2-foss-2022a-long_reads/		* /apps/eb/TSEBRA/1.1.2-foss-2022a-long_reads/
Line 78:		Line 78:
	* /apps/eb/TSEBRA/1.1.2-GCCcore-11.3.0		* /apps/eb/TSEBRA/1.1.2-GCCcore-11.3.0

	~~Be sure to use the version~~ that ~~is compatible with the version of BRAKER~~ you ~~intend to use (e.g.,~~ use the ~~"-long_reads" version of TSEBRA if you're using the "-long_reads" version of BRAKER~~, ~~and <i>don't</i> use the "GCCcore" version of TSEBRA with~~ any ~~of our BRAKER installations as this could cause~~ dependency conflicts or other unforeseeable problems).		It is recommended that you use the second option above, so as to avoid any dependency conflicts or other unforeseeable problems.

	Finally, in order to use BRAKER to predict genes using a protein database, BRAKER needs to use ProtHint; this is included in GeneMark, but BRAKER does not always successfully find it, so you may find it useful to specify the path to it using the following (see also the example submission script below):		Finally, in order to use BRAKER to predict genes using a protein database, BRAKER needs to use ProtHint; this is included in GeneMark, but BRAKER does not always successfully find it, so you may find it useful to specify the path to it using the following (see also the example submission script below):

Jordan at 17:55, 13 June 2024

2024-06-13T17:55:21Z

← Older revision		Revision as of 13:55, 13 June 2024
Line 78:		Line 78:
	* /apps/eb/TSEBRA/1.1.2-GCCcore-11.3.0		* /apps/eb/TSEBRA/1.1.2-GCCcore-11.3.0

	Be sure to use the version that is compatible with the version of BRAKER you intend to use (e.g., use the "-long_reads" version of TSEBRA if you're using the "-long_reads" version of BRAKER, and <i>don't</i> use the "GCCcore" version of TSEBRA with any of our BRAKER installations as this ~~will~~ cause dependency conflicts).		Be sure to use the version that is compatible with the version of BRAKER you intend to use (e.g., use the "-long_reads" version of TSEBRA if you're using the "-long_reads" version of BRAKER, and <i>don't</i> use the "GCCcore" version of TSEBRA with any of our BRAKER installations as this could cause dependency conflicts or other unforeseeable problems).

	Finally, in order to use BRAKER to predict genes using a protein database, BRAKER needs to use ProtHint; this is included in GeneMark, but BRAKER does not always successfully find it, so you may find it useful to specify the path to it using the following (see also the example submission script below):		Finally, in order to use BRAKER to predict genes using a protein database, BRAKER needs to use ProtHint; this is included in GeneMark, but BRAKER does not always successfully find it, so you may find it useful to specify the path to it using the following (see also the example submission script below):