Hi all, I am using bwa mem in this pipeline
bwa mem -t 12 -M -B 4 ref.fas xxx.r1.fastq.gz xxx.r1.fastq.gz > xxx.sam
samtools view -h -b -S -@ 11 xxx.sam > xxx.bam
samtools sort xxx.bam > xxx.sorted.bam
samtools depth xxx.sorted.bam > xxx.bed
Then with R, I read the bed file as dataframe and I plot the data. I get very low depth and a plot that is really fluctuating.
does anybody know if there is a way to improve the depth? thanks a lot for any help.
1 answer
The plot looks fine to me. What read depth are you expecting? From the graph, it appears to have a read depth of 100+, I think that is pretty decent. On the question of peaks, there will always be regions in the genome which will over amplify in PCR amplification (hence sequenced multiple times) step of library preparation, other reason could be that they represent repetitive regions. Have you considered doing a PCR duplicate removal using PICARD, Some of these peaks may disappear.
Log in to answer this question.
-
Difference in depth of coverage between SNIPPY and in-house script
written by Sudheer Gupta •
-
Trying to write bwa mem -> samtools view -> samtools sort loop
written by Lee •
-
No @hd header returned in sam file when running bwa mem
written by MboiTui •
-
BWA alignment/Samtools; Fail to read the header
written by Qboy •
-
I can't see my bam file on igv
written by manubiomed20 •
-
R1 and R2 matching
written by ieie •
-
bwa-mem, from .sam/.bam to fasta file
written by ieie •
-
bwa samtool giving empty bam files while piping
written by Goku •
-
bwa mem algorithm filter secondary alignment
written by xudongyang •
-
Bwa-mem alignment on duplicated regions
written by annecarol7 •
Hello,
you can convert sam to bam and sort on the fly:
What do you mean by "very low depth"?
fin swimmer
when I check the depth at each position, how many times a base is covered as I understood, I see very low depth like from 20 to 300. I thought that the more the depth the more confidence I get in my data. Plus, the plot is really fluctuating while I would expect a more stable plot without many peaks.
Have you done any quality control on your reads?
Have you considered showing us that plot?
this is the plot that I get. thanks for any help!