Buffer overflow when aligning paired-end reads with BWA

[etwatson]

Greetings folks. I'm trying an alignment of paired-end reads with bwa, and the program is crashing with a buffer overflow error. Here is the readout:

...
...
[bwa_sai2sam_pe_core] print alignments... *** buffer overflow detected ***: bwa terminated
======= Backtrace: =========
/lib/x86_64-linux-gnu/libc.so.6(__fortify_fail+0x37)[0x7f9488d954f7]
/lib/x86_64-linux-gnu/libc.so.6(+0xf9410)[0x7f9488d94410]
/lib/x86_64-linux-gnu/libc.so.6(+0xf82e3)[0x7f9488d932e3]
bwa[0x4171bf]
bwa[0x417755]
/lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xed)[0x7f9488cbc30d]
bwa[0x401e21]
======= Memory map: ========
00400000-0042e000 r-xp 00000000 08:01 12191691 /usr/bin/bwa
0062d000-0062e000 r--p 0002d000 08:01 12191691 /usr/bin/bwa
0062e000-00631000 rw-p 0002e000 08:01 12191691 /usr/bin/bwa
020d8000-113a3000 rw-p 00000000 00:00 0 [heap]
7f9484283000-7f9484298000 r-xp 00000000 08:01 12848813 /lib/x86_64-linux-gnu/libgcc_s.so.1
7f9484298000-7f9484497000 ---p 00015000 08:01 12848813 /lib/x86_64-linux-gnu/libgcc_s.so.1
7f9484497000-7f9484498000 r--p 00014000 08:01 12848813 /lib/x86_64-linux-gnu/libgcc_s.so.1
7f9484498000-7f9484499000 rw-p 00015000 08:01 12848813 /lib/x86_64-linux-gnu/libgcc_s.so.1
7f9484499000-7f9488c9b000 rw-p 00000000 00:00 0
7f9488c9b000-7f9488e32000 r-xp 00000000 08:01 12845199 /lib/x86_64-linux-gnu/libc-2.13.so
7f9488e32000-7f9489031000 ---p 00197000 08:01 12845199 /lib/x86_64-linux-gnu/libc-2.13.so
7f9489031000-7f9489035000 r--p 00196000 08:01 12845199 /lib/x86_64-linux-gnu/libc-2.13.so
7f9489035000-7f9489036000 rw-p 0019a000 08:01 12845199 /lib/x86_64-linux-gnu/libc-2.13.so
7f9489036000-7f948903c000 rw-p 00000000 00:00 0
7f948903c000-7f9489054000 r-xp 00000000 08:01 12845213 /lib/x86_64-linux-gnu/libpthread-2.13.so
7f9489054000-7f9489253000 ---p 00018000 08:01 12845213 /lib/x86_64-linux-gnu/libpthread-2.13.so
7f9489253000-7f9489254000 r--p 00017000 08:01 12845213 /lib/x86_64-linux-gnu/libpthread-2.13.so
7f9489254000-7f9489255000 rw-p 00018000 08:01 12845213 /lib/x86_64-linux-gnu/libpthread-2.13.so
7f9489255000-7f9489259000 rw-p 00000000 00:00 0
7f9489259000-7f9489270000 r-xp 00000000 08:01 12848875 /lib/x86_64-linux-gnu/libz.so.1.2.3.4
7f9489270000-7f948946f000 ---p 00017000 08:01 12848875 /lib/x86_64-linux-gnu/libz.so.1.2.3.4
7f948946f000-7f9489470000 r--p 00016000 08:01 12848875 /lib/x86_64-linux-gnu/libz.so.1.2.3.4
7f9489470000-7f9489471000 rw-p 00017000 08:01 12848875 /lib/x86_64-linux-gnu/libz.so.1.2.3.4
7f9489471000-7f94894f4000 r-xp 00000000 08:01 12845203 /lib/x86_64-linux-gnu/libm-2.13.so
7f94894f4000-7f94896f3000 ---p 00083000 08:01 12845203 /lib/x86_64-linux-gnu/libm-2.13.so
7f94896f3000-7f94896f4000 r--p 00082000 08:01 12845203 /lib/x86_64-linux-gnu/libm-2.13.so
7f94896f4000-7f94896f5000 rw-p 00083000 08:01 12845203 /lib/x86_64-linux-gnu/libm-2.13.so
7f94896f5000-7f9489716000 r-xp 00000000 08:01 12845068 /lib/x86_64-linux-gnu/ld-2.13.so
7f94898ec000-7f94898f0000 rw-p 00000000 00:00 0
7f948990d000-7f9489915000 rw-p 00000000 00:00 0
7f9489915000-7f9489916000 r--p 00020000 08:01 12845068 /lib/x86_64-linux-gnu/ld-2.13.so
7f9489916000-7f9489918000 rw-p 00021000 08:01 12845068 /lib/x86_64-linux-gnu/ld-2.13.so
7fff6f097000-7fff6f0b8000 rw-p 00000000 00:00 0 [stack]
7fff6f1ce000-7fff6f1cf000 r-xp 00000000 00:00 0 [vdso]
ffffffffff600000-ffffffffff601000 r-xp 00000000 00:00 0 [vsyscall]
Aborted

I'm not sure what this means, and I hope to gain some degree of enlightenment here. Does this sound like a hardware issue? I am running a Ubuntu machine (11.10) with 70G of memory, and I am able to run this on 1 of 16 threads (Intel Xeon CPU [email protected]).

[Richard Finney]

What version of BWA are you using?
Did you compile it?
Is your index created with the same version of BWA?
You've got a 70GB machine, how much memory is actually available?
What is the command line you're using?
Is your input good?
How soon into the program does it overflow? Any output before the crash?

[etwatson]

1.) Version: 0.5.9-r16
2.) Yes
3.) Yes
4.) see image showing bwa start and crash (I'll have to upload this later)
5.) bwa sampe <ref.fasta> <46sa1.sai> <46sa2.sai> <46sa1.fastq> <46sa2.fastq> > 46sa.sam
6.) not sure. .sai files were created with bwa aln -B 12 -t 10 <index.fa> <file1.fastq>
The fastq files and reference have been used in CLC Bio.
7.) It crashs after 30 seconds.

@PG ID:bwa PN:bwa VN:0.5.9-r16
[bwa_sai2sam_pe_core] convert to sequence coordinate...
[infer_isize] (25, 50, 75) percentile: (151, 182, 228)
[infer_isize] low and high boundaries: 89 and 382 for estimating avg and std
[infer_isize] inferred external isize from 87582 pairs: 191.476 +/- 55.490
[infer_isize] skewness: 0.600; kurtosis: -0.140; ap_prior: 6.11e-05
[infer_isize] inferred maximum insert size: 556 (6.57 sigma)
[bwa_sai2sam_pe_core] time elapses: 10.17 sec
[bwa_sai2sam_pe_core] changing coordinates of 28020 alignments.
[bwa_sai2sam_pe_core] align unmapped mate...
[bwa_paired_sw] 22938 out of 27000 Q17 singletons are mated.
[bwa_paired_sw] 4369 out of 12567 Q17 discordant pairs are fixed.
[bwa_sai2sam_pe_core] time elapses: 14.51 sec
[bwa_sai2sam_pe_core] refine gapped alignments... 1.94 sec
[bwa_sai2sam_pe_core] print alignments... *** buffer overflow detected ***: bwa terminated

[etwatson]

Greetings all, I'm guessing that nobody else has encountered this error. Can anyone help me trouble shoot it? What is a buffer overflow, and what can I do to get bwa working?