wangyugui

@pd3 Yet no official benchmarks, but there are some cases. case1: In a pipline such as 'bwa | samtools fixmate | samtools sort |samtools markdup', we notice that sometime 'crc'...

And we need 'wbu' mode for btrfs transparent compress too?

we need both solution for these two cases 1) fast access for intermediate file. 2) small space for final file. ssd/nvme become faster, and more cpu/core is coming, we need...

maybe 0.002 will be better than 20K CPM. if we use 0.002 , then we will not need a new param interface.

we can set different --max-bundle-frags for every bam when cufflinks. but we have to set one value for cuffdiff with multiple bam file.

I worked around this problem by mask the bam file by samtools. ``` samtools view -L ${ref_dir}/cufflinks.mask.bed -u -o mask.bam -U cufflink.bam my.bam ```

``` java -jar /usr/hpc-bio/gatk/gatk-package-4.0.3.0-local.jar HaplotypeCaller -L /biowrk/BaseSpace/vcf.NA12878/vcf.gatk4.bwa.HiSeqX-PCR-free-v2.5-70x/tmp_dir/0000-scattered.intervals -O /biowrk/BaseSpace/vcf.NA12878/vcf.gatk4.bwa.HiSeqX-PCR-free-v2.5-70x/tmp_dir/gatk.1.vcf --arguments_file /biowrk/BaseSpace/vcf.NA12878/vcf.gatk4.bwa.HiSeqX-PCR-free-v2.5-70x/gatk.arguments_file.txt --verbosity WARNING --reference /usr/bio-ref/GRCh38.p0.dnaref/dnaref.fa --dbsnp /usr/bio-ref/GRCh38.p0.dnaref/dbsnp.vcf --max-reads-per-alignment-start 100 # cat /biowrk/BaseSpace/vcf.NA12878/vcf.gatk4.bwa.HiSeqX-PCR-free-v2.5-70x/gatk.arguments_file.txt -I /biowrk/BaseSpace/bam.bwa/HiSeqX-PCR-free-v2.5-NA12878.70x/md.bam ```

we need to detect the exist filesystem and ' -F, --force' to overwrite it?

a dirty fix to wipefs when 'mkfs.ntfs --fast' ``` @@ -3539,6 +3539,12 @@ static BOOL mkntfs_open_partition(ntfs_volume *vol) } ntfs_log_warning("mkntfs forced anyway. Hope /etc/mtab is incorrect.\n"); } + { + static...