search

BSSeeker / BSseeker2

A versatile aligning pipeline for bisulfite sequencing data
http://pellegrini.mcdb.ucla.edu/BS_Seeker2/
MIT License
60 stars 25 forks source link

[HELP] The output of bs_seeker2-call_methylation.py #44

Open hmyh1202 opened 2 years ago

hmyh1202 commented 2 years ago

Hello;

I have tested the BSseeker2 for my data, and run the Bismark and BSMAP for the same data. The parameters of BSMAP and Bismark was default. The align step of BSseeker is default, and the parameters of mC calling step is bs_seeker2-call_methylation.py -i sort.dedup.bam -o test --db bsseeker2/wgbs/tair10.fa_bowtie2/ --rm-overlap. And the total cytosine that detedted is 6,547,483 for BSbseeker2, 42,486,818 for Bismark and 41,535,940 for BSMAP.

The mapping ratio is 80.25% for bismark, 72.10% for BSMAP and 76.74% for BSseeker2. I did not notice any error in alignment step and deduplication step.

So the cytosine number of BSseeker is so low. How can I handle this issue.

Thank you !

guoweilong commented 2 years ago

You can check the CGmap file to see if all the chromosomes were correctly covered. And it would also be easier to see what would be the sites not found in the results as you calculate using BS-Seeker2 but by the others? Would they be a matter of the coverage threshold or '--rm-overlap'?

Weilong

At 2022-04-12 17:28:04, "Grace Bisulfite" @.***> wrote:

Hello;

I have tested the BSseeker2 for my data, and run the Bismark and BSMAP for the same data. The parameters of BSMAP and Bismark was default. The align step of BSseeker is default, and the parameters of mC calling step is bs_seeker2-call_methylation.py -i sort.dedup.bam -o test --db bsseeker2/wgbs/tair10.fa_bowtie2/ --rm-overlap. And the total cytosine that detedted is 6,547,483 for BSbseeker2, 42,486,818 for Bismark and 41,535,940 for BSMAP.

The mapping ratio is 80.25% for bismark, 72.10% for BSMAP and 76.74% for BSseeker2. I did not notice any error in alignment step and deduplication step.

So the cytosine number of BSseeker is so low. How can I handle this issue.

Thank you !

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you are subscribed to this thread.Message ID: @.***>

hmyh1202 commented 2 years ago

I hava had checked the CGmap file, and all the chromosomes were correctly covered. The results showed that the least number of reads covering one site is 1x.

And I do not find parameters that who can adjust the coverage threshold, expect for --read-no for wig output.

Also, I remove the --rm-overlap, and the output of the cytosine number covered by 1x is 6,465,486, not increase.

Thank you.

guoweilong commented 2 years ago

It's wired. Did you observe significant difference in the read counts in bam files those were used for call-methylation step by the three aligners?

Weilong

At 2022-04-13 09:31:44, "Grace Bisulfite" @.***> wrote:

I hava had checked the CGmap file, and all the chromosomes were correctly covered. The results showed that the least number of reads covering one site is 1x.

And I do not find parameters that who can adjust the coverage threshold, expect for --read-no for wig output.

Also, I remove the --rm-overlap, and the output of the cytosine number covered by 1x is 6,465,486, not increase.

Thank you.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: @.***>