search

arq5x / lumpy-sv

lumpy: a general probabilistic framework for structural variant discovery
MIT License
307 stars 119 forks source link

Lumpy_filter error and which is the best way to generate the discordant paired-end alignments #335

Open biozzq opened 4 years ago

biozzq commented 4 years ago

Dear @brentp

Recently, I tried to use smoove to detect CNV using the BAM file generated by BWA-MEM as input. However, lumpy_filter always gave me an error: "Segmentation fault (core dumped)"; thus, I changed to use lumpy tranditional usage. During generating the discordant paired-end alignments, I found that the results from the following two pipelines are very different. In my test, the samblaster just reported 0 discordant read pair, but samtools reported many. (1) samtools view -b -F 1294 sample.bam (2) samtools sort -@ 4 -O SAM -n $bam | samblaster -a -e -d $prefix.disc.sam -s $prefix.split.sam -o /dev/null

Hope you can help me with above problems and looking forward to your reply.

Sincerely, Zheng Zhuqing

biozzq commented 4 years ago

Dear all,

I think we can refer to this link (https://github.com/GregoryFaust/samblaster/issues/46), hope this can help you.

Sincerely, Zhuqing

biozzq commented 4 years ago

Dear @brentp

While the tranditional usage is ok now, the smoove always failed at lumpy_filter and only gave me limited error information as described above. I do not know why. Hope your help.

sincerely， Zhuqing

brentp commented 4 years ago

if you have an issue with smoove, you can open an issue there with full details (the command you ran along with full STDOUT and STDERR).

GregoryFaust commented 4 years ago

GregoryFaust/samblaster#46 is closed in https://github.com/GregoryFaust/samblaster/releases/tag/v.0.1.26