search

blahah / transrate

Understand your transcriptome assembly
http://hibberdlab.com/transrate
Other
100 stars 34 forks source link

SNAP failing in v 1.0.3 #218

Open ssander5 opened 6 years ago

ssander5 commented 6 years ago

I'm running transrate with the --left and --right options. The inputs are correct, full fq files. I've run it on up to 300GB of memory. I've reinstalled transrate using the source from a previous issue with the same problem - I'm still getting the error:

Welcome to SNAP version 1.0dev.96. mmap failed. MemMapDataReader: fail to map /N/dc2/scratch/hewals/BACKUP/input_files/Cat_R1.fastq at 0,107772424313 SNAP exited with exit code 1 from line 2283 of file SNAPLib/DataReader.cpp

I'm using this command: transrate --assembly /N/dc2/scratch/hewals/BACKUP/final_assemblies/okayset/combined.okay.fa --left=/N/dc2/scratch/hewals/BACKUP/input_files/Cat_R1.fastq --right=/N/dc2/scratch/hewals/BACKUP/input_files/Cat_R2.fastq

and I'm running this on a high memory cluster. I have run into this error on every dataset I have tested. Any advise would be appreciated!

mscheremetjew commented 6 years ago

Hi, Have you tried to install the reference-based mode dependencies? This solved the issue I had with SNAP on my side. Command: transrate --install-deps ref

isgilman commented 5 years ago

I'm having a similar and possibly related problem. Here is my command and output.

[isg4@c13n05 Transrate]$ transrate --assembly ERR204.Trinity.fasta --left ERR2040261_1.overep_filtered.fq --right ERR2040261_2.overep_filtered.fq --threads 20 --output Transrate_output/

[ INFO] 2019-03-06 16:05:56 : Loading assembly: ERR204.Trinity.fasta
[ INFO] 2019-03-06 16:06:01 : Analysing assembly: ERR204.Trinity.fasta
[ INFO] 2019-03-06 16:06:01 : Results will be saved in Transrate_output/ERR204.Trinity
[ INFO] 2019-03-06 16:06:01 : Calculating contig metrics...
[ INFO] 2019-03-06 16:06:05 : Contig metrics:
[ INFO] 2019-03-06 16:06:05 : -----------------------------------
[ INFO] 2019-03-06 16:06:05 : n seqs                        37166
[ INFO] 2019-03-06 16:06:05 : smallest                        201
[ INFO] 2019-03-06 16:06:05 : largest                        9434
[ INFO] 2019-03-06 16:06:05 : n bases                    30355555
[ INFO] 2019-03-06 16:06:05 : mean len                     816.76
[ INFO] 2019-03-06 16:06:05 : n under 200                       0
[ INFO] 2019-03-06 16:06:05 : n over 1k                     10373
[ INFO] 2019-03-06 16:06:05 : n over 10k                        0
[ INFO] 2019-03-06 16:06:05 : n with orf                    14903
[ INFO] 2019-03-06 16:06:05 : mean orf percent              67.49
[ INFO] 2019-03-06 16:06:05 : n90                             346
[ INFO] 2019-03-06 16:06:05 : n70                             751
[ INFO] 2019-03-06 16:06:05 : n50                            1222
[ INFO] 2019-03-06 16:06:05 : n30                            1774
[ INFO] 2019-03-06 16:06:05 : n10                            2720
[ INFO] 2019-03-06 16:06:05 : gc                             0.44
[ INFO] 2019-03-06 16:06:05 : bases n                           0
[ INFO] 2019-03-06 16:06:05 : proportion n                    0.0
[ INFO] 2019-03-06 16:06:05 : Contig metrics done in 4 seconds
[ INFO] 2019-03-06 16:06:05 : Calculating read diagnostics...
[ERROR] 2019-03-06 16:06:14 : Failed to build Snap index
Welcome to SNAP version 1.0dev.96.

Saving genome...0s

Using print statements I figured out that the "Failed to build Snap index" error is being thrown from line 146 (msg = "Failed to build Snap index\n#{runner.stderr}" instead of line 127, which has the same error message) of snap.rb.

Solutions I've tried (to no avail):

lmuenter commented 3 years ago

On a side note: You can download the Oyster River Protocol's packaged version of Transrate. It's standalone and works like a charm. It's sad that this program has been discontinued!

BioinfoAK commented 1 year ago

Hi @isgilman, did you ever get around to overcoming the error you were getting? I'm in the same boat and have inserted the cmd << " -mcp 10000000" # maximum candidate pool size fix already.

isgilman commented 1 year ago

Sorry @BioinfoAK, this was so long ago I can't quite remember what steps I took afterwards, but I don't think I continued with transrate. Good luck!