core dumped happened when when mapping against transcriptome.

[Yugui Wang]

Hi.

core dumped happened when when mapping against transcriptome.


#STAR --version
STAR_2.5.2a

# STAR --runThreadN 48 --outSAMtype SAM --genomeSAindexNbases 15 --outSAMunmapped None --genomeDir ../database/ --readFilesIn /biowrk/fastq.clean/Project_14688/Sample_118/_1.fq /biowrk/fastq.clean/Project_14688/Sample_118/_2.fq --outFileNamePrefix ./ --twopassMode Basic --chimSegmentMin 25 --outSAMstrandField intronMotif --outSAMattributes NH HI AS nM NM --outFilterIntronMotifs RemoveNoncanonical --chimOutType SeparateSAMold --outFilterMultimapNmax 2
Jun 06 00:48:32 ..... started STAR run
Jun 06 00:48:32 ..... loading genome
Jun 06 00:48:36 ..... started 1st pass mapping
Segmentation fault      (core dumped)

(gdb) where
#0  operator[] (ii=18446744073709551615, this=<optimized out>) at PackedArray.h:30
#1  compareSeqToGenome (s2=0x50a1280, S=35, N=36, L=15, g=0x7f2c8efc50d8 "", SA=..., iSA=18446744073709551615, dirR=false,
    compRes=@0x7f2939d80440: true, P=0x16c5a80) at SuffixArrayFuns.cpp:17
#2  0x000000000040bffd in maxMappableLength (s=0x50a1280, S=S@entry=35, N=N@entry=36, g=0x7f2c8efc50d8 "", SA=..., i1=0,
    i2=18446744073709551615, dirR=dirR@entry=false, L=@0x7f2939d80590: 15, indStartEnd=indStartEnd@entry=0x7f2939d805c0,
    P=P@entry=0x16c5a80) at SuffixArrayFuns.cpp:139
#3  0x00000000004486cf in ReadAlign::maxMappableLength2strands (this=this@entry=0x7f2aaab31010, pieceStartIn=pieceStartIn@entry=35,
    pieceLengthIn=36, iDir=iDir@entry=1, iSA1=iSA1@entry=0, iSA2=<optimized out>, maxLbest=@0x7f2939d807c0: 0, iFrag=0)
    at ReadAlign_maxMappableLength2strands.cpp:82
#4  0x000000000043cbfa in ReadAlign::mapOneRead (this=this@entry=0x7f2aaab31010) at ReadAlign_mapOneRead.cpp:59
#5  0x000000000044e90e in ReadAlign::oneRead (this=0x7f2aaab31010) at ReadAlign_oneRead.cpp:70
#6  0x0000000000443cb4 in ReadAlignChunk::mapChunk (this=this@entry=0x34668f0) at ReadAlignChunk_mapChunk.cpp:25
#7  0x0000000000441f4e in ReadAlignChunk::processChunks (this=0x34668f0) at ReadAlignChunk_processChunks.cpp:144
#8  0x000000000047a5c9 in ThreadControl::threadRAprocessChunks (RAchunk=<optimized out>) at ThreadControl.h:22
#9  0x00007f2d7c27fdc5 in start_thread () from /lib64/libpthread.so.0
#10 0x00007f2d7bfacced in clone () from /lib64/libc.so.6

[Alexander Dobin]

Hi Yugui,

please send me the Log.out file of the genome generation step. If your reference is very small, you need to reduce --genomeSAindexNbases  - it has to be scaled with the genome length, as ~min(14,log2(ReferenceLength)/2 - 1).

Cheers

Alex

[Yugui Wang]

Hi.

Please see the attachment file(Log.out/--genomeSAindexNbases 15).

I will reduce reduce --genomeSAindexNbases to 12 , and test it again.


在 2016年6月8日星期三 UTC+8上午5:25:01，Alexander Dobin写道：

[Yugui Wang]

The STAR with --genomeSAindexNbases 12  sucessfully finished.

Although there is some info in doc.
For small genomes, the parameter --genomeSAindexNbases must to be scaled down, with a typical
value of min(14, log2(GenomeLength)/2 - 1).

But there is also some info in doc.
Longer strings will use much more memory, but allow faster searches.

Can we use bigger genomeSAindexNbases for small genome when we have enough memory,
and we want faster search?

Best Regards
Wang Yugui




在 2016年6月12日星期日 UTC+8下午7:00:40，Yugui Wang写道：

[Alexander Dobin]

Hi Yugui,

this is impossible at the moment because of the seg-fault.

For small genomes, the speed advantage will be minimal if any.

Cheers

Alex

[Yugui Wang]

Although STAR with --genomeSAindexNbases 12 is OK when mapping to human transcriptome(400M).

But both --genomeSAindexNbases 12/11/10/9/8/7 seg-faulted when mapping to human HLA transcriptome(8.5M).

more detail please see the attachment file(Log.out when --genomeSAindexNbases 11)


+ STAR --runMode genomeGenerate --runThreadN 48 --limitGenomeGenerateRAM 240000000000 --genomeSAindexNbases 11 --genomeFastaFiles ../database/STAR.fa --genomeDir ../database/
Jun 16 19:42:07 ..... started STAR run
Jun 16 19:42:07 ... starting to generate Genome files
Jun 16 19:43:02 ... starting to sort Suffix Array. This may take a long time...
Jun 16 19:43:13 ... sorting Suffix Array chunks and saving them to disk...
Jun 16 19:43:45 ... loading chunks from disk, packing SA...
Jun 16 19:43:50 ... finished generating suffix array
Jun 16 19:43:50 ... generating Suffix Array index
Jun 16 19:43:51 ... completed Suffix Array index
Jun 16 19:43:51 ... writing Genome to disk ...
Jun 16 19:43:54 ... writing Suffix Array to disk ...
Jun 16 19:43:54 ... writing SAindex to disk
Jun 16 19:43:55 ..... finished successfully
+ unstranded_options='--outSAMstrandField intronMotif --outSAMattributes NH HI AS nM NM'
+ cufflinks_options='--outFilterIntronMotifs RemoveNoncanonical'
+ STAR --runThreadN 48 --outSAMtype SAM --genomeSAindexNbases 11 --outSAMunmapped None --genomeDir ../database/ --readFilesIn /biowrk/fastq.clean/Project_14688/Sample_118/_1.fq /biowrk/fastq.clean/Project_14688/Sample_118/_2.fq --outFileNamePrefix ./ --twopassMode Basic --chimSegmentMin 25 --outSAMstrandField intronMotif --outSAMattributes NH HI AS nM NM --outFilterIntronMotifs RemoveNoncanonical --chimOutType SeparateSAMold --outFilterMultimapNmax 6
Jun 16 19:43:55 ..... started STAR run
Jun 16 19:43:55 ..... loading genome
Jun 16 19:43:57 ..... started 1st pass mapping
./run.sh: line 52: 80840 Segmentation fault      (core dumped) STAR --runThreadN ${NPROC} --outSAMtype SAM --genomeSAindexNbases 11 --outSAMunmapped None --genomeDir ../database/ --readFilesIn /biowrk/fastq.clean/Project_14688/Sample_118/_1.fq /biowrk/fastq.clean/Project_14688/Sample_118/_2.fq --outFileNamePrefix ./ --twopassMode Basic --chimSegmentMin 25 ${unstranded_options} ${cufflinks_options} --chimOutType SeparateSAMold --outFilterMultimapNmax 6

[Alexander Dobin]

Hi Yugui

could you please send me the link to the "HLA" reference?

Cheers

Alex