a_to_g mode seg faults

[rei...@gmail.com]

I'm trying to align single end RNAseq reads in a_to_g mode to find RNA editing sites, but the new version GNUMAP 4.0.0, keeps giving me a segmentation fault error.

It seems to be specific to a_to_g mode, since it works fine on the same test dataset without -d, but seg faults if -d is included. See below:

Here is a run with a_to_g off:

[u0752512@ember2 gnu_test]$ /uufs/chpc.utah.edu/common/home/u0752512/gnumap-bwt/bin/gnumap -g /uufs/chpc.utah.edu/common/home/u0752512/dmel/dmel-all-chromosome-r6.21.fasta -j 20 -m 12 -c 10 -v 1 -o ./wt_test wt_test.fastq

This is GNUMAP, Version 4.0.0 BETA, for public and private use.

# Using BWT/Suffix Array.

Command Line Arguments:  /uufs/chpc.utah.edu/common/home/u0752512/gnumap-bwt/bin/gnumap -g /uufs/chpc.utah.edu/common/home/u0752512/dmel/dmel-all-chromosome-r6.21.fasta -j 20 -m 12 -c 10 -v 1 -o ./wt_test wt_test.fastq 

Parameters: 

Verbose: 1

Genome file(s): /uufs/chpc.utah.edu/common/home/u0752512/dmel/dmel-all-chromosome-r6.21.fasta

Output file: ./wt_test

Max matches: 1000

Sequence file(s): wt_test.fastq

Align percentage: 90%

Using new phred quality scores.

Using Needleman-Wunsch alignments

Number of threads: 10

Mer size: 12

Using jump size of 20

Using min seed matches for each sequence of 2

Using Default Alignment Scores

Gap score: -1

Maximum Gaps: 3

Indexing the genome.

Genome reference already exists and loaded successfully!

Success!

Time to index: 0 seconds

Matching 1 file(s):

[-/0] Matching 125000 sequences of: wt_test.fastq

Reads per processor: 2048

[0/0] 41% reads complete

[0/0] 57% reads complete

[0/0] 77% reads complete

[0/0] 97% reads complete

[0/0] 100% reads complete

[0/0] 100% reads complete

Total NW: 33841, Max NW: 33841, Min NW: 33841, DIFF: 0

[0/-] Time since start: 30.0396

[0/-] Printing output.

Finished printing to ./wt_test.sgr

#Finished.

# Total Time: 30.6097 seconds.

# Time spent waiting: 0.437485 + 0.183944=0.621428 seconds.

# Found 125000 sequences.

# Sequences matched: 108677

# Sequences not matched: 16323

# Output written to ./wt_test.sam

Total wait time is 2.447395

And the same run with a_to_g turned on:

[u0752512@ember2 gnu_test]$ /uufs/chpc.utah.edu/common/home/u0752512/gnumap-bwt/bin/gnumap -g /uufs/chpc.utah.edu/common/home/u0752512/dmel/dmel-all-chromosome-r6.21.fasta -j 20 -m 12 -c 10 -v 1 -d -o ./wt_test wt_test.fastq

This is GNUMAP, Version 4.0.0 BETA, for public and private use.

# Using BWT/Suffix Array.

Command Line Arguments:  /uufs/chpc.utah.edu/common/home/u0752512/gnumap-bwt/bin/gnumap -g /uufs/chpc.utah.edu/common/home/u0752512/dmel/dmel-all-chromosome-r6.21.fasta -j 20 -m 12 -c 10 -v 1 -d -o ./wt_test wt_test.fastq 

Parameters: 

Verbose: 1

Genome file(s): /uufs/chpc.utah.edu/common/home/u0752512/dmel/dmel-all-chromosome-r6.21.fasta

Output file: ./wt_test

Max matches: 1000

Sequence file(s): wt_test.fastq

Align percentage: 90%

Using new phred quality scores.

Using Needleman-Wunsch alignments

Number of threads: 10

Mer size: 12

Using jump size of 20

Using min seed matches for each sequence of 2

Using Default Alignment Scores

Gap score: -1

Maximum Gaps: 3

Using A to G conversion

Using irregular bin size of 1

Indexing the genome.

Genome reference already exists and loaded successfully!

Using AtoG conversion techniques

Success!

Time to index: 0 seconds

Matching 1 file(s):

[-/0] Matching 125000 sequences of: wt_test.fastq

Reads per processor: 2048

Segmentation fault

[u0752512@ember2 gnu_test]$ 

[M. Stanley Fujimoto]

Thanks for using the software! I'll take a look and let you know what I find.

[M. Stanley Fujimoto]

I think I've got the bug fixed. Sorry for any inconvenience this has caused you.

We've pushed a 4.0.1 release here https://github.com/byucsl/gnumap/releases

Let me know if it works for you.