memory error

[xfliu1995]

Too much memory is occupied by software operation, when mapping by STAR. This often causes software to fail.How do you limit the maximum amount of memory that your software can use while it is running?

[shapemapper]

What are your alignment target sequences?

If you're attempting to align against a large number of transcripts, see FAQ. On the other hand, if you're trying to align against a small number of large sequences (for example the human genome), the aligner index will just require a lot of memory (~30Gb in the case of the human genome), and that's not easily worked around without changing the aligner.

[shapemapper]

I should also add that you can adjust the speed/memory tradeoff with the --genomeSAindexNbase parameter, which is passed through to STAR during the genome index building phase.

[xfliu1995]

When I add --genomeSAindexNbase parameter for shapemapper2，shapemapper2 running error.
Traceback (most recent call last):
File "/BioII/lulab_b/shibinbin/projects/shapemap/python/cli.py", line 108, in
run(sys.argv)
File "/BioII/lulab_b/shibinbin/projects/shapemap/python/cli.py", line 46, in run
pipeline = ap.construct(rest_args)
File "/BioII/lulab_b/shibinbin/projects/shapemap/python/pyshapemap/pipeline_arg_parser.py", line 595, in construct
params, fastqs = parse_args(args)
File "/BioII/lulab_b/shibinbin/projects/shapemap/python/pyshapemap/pipeline_arg_parser.py", line 440, in parse_args
raise RuntimeError("Error: unrecognized argument(s): {}".format(rest))
RuntimeError: Error: unrecognized argument(s): ['--genomeSAindexNbases']

[shapemapper]

Are you running the most recent version of shapemapper?