Error running abra [0.96]

[chisqr]

Hello,

I was able to run abra for a number of BAM files. But a few of them are giving me this error:

-- Fri Nov 13 03:44:57 CST 2015 : Indexing contigs Running: [bwa index abra_temp_dir/clean_contigs.fasta] [bwa_index] Pack FASTA... 0.21 sec [bwa_index] Construct BWT for the packed sequence... [bwa_index] 7.06 seconds elapse. [bwa_index] Update BWT... 0.13 sec [bwa_index] Pack forward-only FASTA... 0.15 sec [bwa_index] Construct SA from BWT and Occ... 2.10 sec [main] Version: 0.7.8-r455 [main] CMD: bwa index abra_temp_dir/clean_contigs.fasta [main] Real time: 10.393 sec; CPU: 9.660 sec Stream thread done. Stream thread done. BWA time: 10 seconds. Fri Nov 13 03:45:08 CST 2015 : Contig indexing done Fri Nov 13 03:45:08 CST 2015 : Aligning original reads to contigs Running: [bwa aln abra_temp_dir/clean_contigs.fasta abra_temp_dir/temp1/original_reads.fastq.gz -t 12 -o 0 | bwa samse abra_temp_dir/clean_contigs.fasta - abra_temp_dir/temp1/original_reads.fastq.gz -n 1000] #

A fatal error has been detected by the Java Runtime Environment:

#

SIGSEGV (0xb) at pc=0x00007fdc8fd785fe, pid=6272, tid=140585289746176

#

JRE version: Java(TM) SE Runtime Environment (8.0_20-b26) (build 1.8.0_20-b26)

Java VM: Java HotSpot(TM) 64-Bit Server VM (25.20-b23 mixed mode linux-amd64 compressed oops)

Problematic frame:

V [libjvm.so+0x8e35fe] SRhandler(int, siginfo, ucontext_)+0x3e

#

Core dump written. Default location:

#

An error report file with more information is saved as:

--

I am using Version: 0.7.8-r455 of BWA

Can you please help me with this ? Thanks a lot for updating the software !

regards, Rahul

[mozack]

Are you able to provide another BAM snippet that reproduces the problem?

[chisqr]

Here's one more

-- Abra JNI entry point v0.96, prefix: X_140994966_140995366, read_length: 101, kmer_size: 47, min_node_freq: 2, min_base_qual: 60, min_edge_ratio 0.020000, debug: 1, max_nodes: 9000 Assembling: -> abra_temp_dir/X_140994966_140995366_contigs.fasta_k47 STOPPED_ON_REPEAT: X_140994966_140995366 Done assembling(0): abra_temp_dir/X_140994966_140995366_contigs.fasta_k47, 0 Abra JNI entry point v0.96, prefix: X_140994966_140995366, read_length: 101, kmer_size: 49, min_node_freq: 2, min_base_qual: 60, min_edge_ratio 0.020000, debug: 1, max_nodes: 9000 Assembling: -> abra_temp_dir/X_140994966_140995366_contigs.fasta_k49 STOPPED_ON_REPEAT: X_140994966_140995366 Done assembling(0): abra_temp_dir/X_140994966_140995366_contigs.fasta_k49, 0 Abra JNI entry point v0.96, prefix: X_140994966_140995366, read_length: 101, kmer_size: 51, min_node_freq: 2, min_base_qual: 60, min_edge_ratio 0.020000, debug: 1, max_nodes: 9000 Assembling: -> abra_temp_dir/X_140994966_140995366_contigs.fasta_k51 Done assembling(1): abra_temp_dir/X_140994966_140995366_contigs.fasta_k51, 2 Elapsed_msecs_in_NativeAssembler Region: X_140994966_140995366 Length: 400 ReadCount: 460 Elapsed 794 Assembled true 23 Done assembling(3): abra_temp_dir/X_77244911_77245471_contigs.fasta_k17, 1033 Elapsed_msecs_in_NativeAssembler Region: X_77244911_77245471 Length: 560 ReadCount: 1241 Elapsed 2347 Assembled true 17 Clock time in Assembly: 475 Fri Nov 13 03:49:40 CST 2015 : Aligning contigs Running: [bwa mem -t 14 /reference/gatk/bundle/2.8/human_g1k_v37_decoy.fasta abra_temp_dir/all_contigs.fasta > abra_temp_dir/all_contigs.sam] #

A fatal error has been detected by the Java Runtime Environment:

#

SIGSEGV (0xb) at pc=0x00007fd32595b5fe, pid=38053, tid=140544852342528

#

JRE version: Java(TM) SE Runtime Environment (8.0_20-b26) (build 1.8.0_20-b26)

Java VM: Java HotSpot(TM) 64-Bit Server VM (25.20-b23 mixed mode linux-amd64 compressed oops)

Problematic frame:

V [libjvm.so+0x8e35fe] SRhandler(int, siginfo, ucontext_)+0x3e

#

An error report file with more information is saved as:

--

I am using the binaries provided here, should I compile them myself and try ?

[chisqr]

I am running abra with 14 threads and using jdk1.8.0_20

[mozack]

14 threads and java 1.8 should be fine.

The error message, unfortunately is not very informative.

You could try compiling yourself, however the fact that some samples succeed leads me to believe that might not help.