This is a test version of Biostars. For the public version, visit https://www.biostars.org.
percent methylation in CHH file of Bismark?

Hi, I have generated a file by cytosine report function in bismark to calculate even the non CpG methylated Cs(CHH,CHG). The file of CHH shown as follow:

 > chr1 3000001 -   0   0   CHH CNN
    chr1    3000006 +   0   0   CHH CTT
    chr1    3000011 +   0   0   CHH CTA
    chr1    3000015 -   0   0   CHH CAT
    chr1    3000021 -   0   0   CHH CTA
    chr1    3000030 -   0   0   CHH CAT
    chr1    3000038 -   0   0   CHH CCA
    chr1    3000039 -   0   0   CHH CCC
    chr1    3000041 -   0   0   CHH CAC
    chr1    3000054 +   0   0   CHH CTT
    chr1    3000059 -   0   0   CHH CAA
    chr1    3000061 +   0   0   CHH CCT
    chr1    3000062 +   0   0   CHH CTT
    chr1    3000065 +   0   0   CHH CTT
    chr1    3000073 +   0   0   CHH CCT
    chr1    3000074 +   0   0   CHH CTA
    chr1    3000082 +   0   0   CHH CTT
    chr1    3000086 -   0   0   CHH CTA
    chr1    3000087 -   0   0   CHH CCT
    chr1    3000091 -   0   0   CHH CAA
    chr1    3000092 -   0   0   CHH CCA

I have to calculate the total coverage at each location and % methylation and for this the formula i know is 

`column4 of '+' strand + column5 of '+' strand + column4 of '-' strand + column5 of '-' strand]= total coverage`

and percentage was equal to [($4/$4+$5)*100 of '+'strand +($4/$4+$5)*100 of -strand]/2 but this could only b possible in CpG cytosine covergae and CHG coverage files that have output format like:`

chr1    3000035 +   0   0   CHG CTG
chr1    3000037 -   0   0   CHG CAG
chr1    3000045 +   0   0   CHG CAG
chr1    3000047 -   0   0   CHG CTG`

means alternate + and negative strand. But in CHH file there are no alternate strands and even the gaps between the values is not consistent. I dont know how to calculate these values. Should I do it for each single line means total coverage will be then sum of column 4 and 5 and not to bother about + strand or its relative negative strand.Any suggestions

next-gen
• 3,165 views • link
written by ahsan.raza.rana • ahsan.raza.rana

2 answers

For CHH sites you calculate each line, there is no merging of lines because there are no C's on the opposite strand to merge with.

• 2 views • link
written by Devon Ryan Devon Ryan

thanks alot. I am really grateful

• 2 views • link
written by ahsan.raza.rana • ahsan.raza.rana

Log in to answer this question.

More posts like this