Hi all.
I know this has been posted already quite a few times before, but I've been trying most of the suggestions offered here on the forum and other forums and I'm still stuck with the same problem. So, I'm hoping that you guys still have some other ideas to help me out here :-)
This is my problem:
I am getting really low mapping efficiencies (15 to 18%) with my 90 bp paired-end bisulfite sequencing reads using Bismark/Bowtie2.
I have been quality trimming my sequences to get rid of adapter sequences and low quality base pairs using Trim Galore! at its standard settings, which are quite stringent.
FastQC didn't report any problems, except for those that can be expected for bisulfite treated data (ie. k-mer content). So no over-represented or duplicated sequences were detected.
I have been using different Bowtie2 settings, like increasing the sensitivity function, decreasing the mapping penalties through the scoring function, setting the insert size to a range as wide as 0 to 1500 bp... but nothing got me higher mapping efficiencies than 15 to 18%. Ambiguous mapped reads were each time around 8%, which is to be expected considering the fact that the genome I am working with is known to have a quite high number of gene duplications.
In conclusion, I am really running out of options and ideas and I am hoping that you guys might have some more experience with it so I can get these efficiencies up!
Thanks a ton!
I know this has been posted already quite a few times before, but I've been trying most of the suggestions offered here on the forum and other forums and I'm still stuck with the same problem. So, I'm hoping that you guys still have some other ideas to help me out here :-)
This is my problem:
I am getting really low mapping efficiencies (15 to 18%) with my 90 bp paired-end bisulfite sequencing reads using Bismark/Bowtie2.
I have been quality trimming my sequences to get rid of adapter sequences and low quality base pairs using Trim Galore! at its standard settings, which are quite stringent.
FastQC didn't report any problems, except for those that can be expected for bisulfite treated data (ie. k-mer content). So no over-represented or duplicated sequences were detected.
I have been using different Bowtie2 settings, like increasing the sensitivity function, decreasing the mapping penalties through the scoring function, setting the insert size to a range as wide as 0 to 1500 bp... but nothing got me higher mapping efficiencies than 15 to 18%. Ambiguous mapped reads were each time around 8%, which is to be expected considering the fact that the genome I am working with is known to have a quite high number of gene duplications.
In conclusion, I am really running out of options and ideas and I am hoping that you guys might have some more experience with it so I can get these efficiencies up!
Thanks a ton!
Comment