depth of coverage and read depth using Bismark

[Hedi86]

hello everyone

is it possible to calculate depth of coverage and read depth using Bismark ? if yes how and if No what is the best way to calculate depth of coverage and read of depth from RRBS reads taking from illumina sequencing ?

Best

[fkrueger]

Bismark itself is merely the alignment tool, and as such does not do any further analysis. I am sure there are several different options to choose from to assess sequence coverage, I can only recommend SeqMonk which offers visualisation and calculations in one package.

[Hedi86]

thank you for your help.

i tried to use seqmonk, but i wasnt sure which file should i import (.fq file after trimming or COV file from bismark) ? in addition what is the correct pipeline in seqmonk, is it ok to import>define probes>running window generator>quantitation>% coverage quantitation OR Coverage depth quantitation ? Seqmonk then report a summary but in summary report noting showing the coverage or Coverage depth.

thanks again

[fkrueger]

It depends a little on which kind of statistics you are interested in. If you really just want to get a value for a fold-coverage of your experiment you should probably import the (deduplicated) Bismark BAM files into SeqMonk because this statistic is a function of the total read length of all reads in your experiment and the size of the genome.

If you import those BAM files you can do e.g. a running Window probe generation followed by a read count quantitation. If you then Click on "DataStore Summary Report" you will see a report that has - among many fields - a Fold Coverage column: this is the value you are looking for. You can also get a mean/median quantification for your probe of interest.

If you wanted to get similar statistic for single-base resolution cytosines you could import the coverage files, design probes over the Read Positions and count their abundance. But yea, it depends on what you really want to get out in the end.

[Hedi86]

really appreciate your comment

so could we say that 0.04 as fold coverage is equal to 4X depth coverage, and could we say that read coverage is equal to mapping efficiency?

thanks again

[fkrueger]

Not quite. A 0.04 fold-coverage means that - on average - each position in the genome was covered 0.04 times. A 4-fold would be, well, 4.

This value is obviously not so meaningful if you used only methylation calls to calculate the value, since they are only 1bp long ( (total number of reads * read length) / length of genome = fold-coverage).

The mapping efficiency is a mapping specific parameter that has to do with which fraction of a sequencing file comes from mappable parts of the genome, and as such has nothing to do with fold-coverage. If you take a library with a 70% mapping efficiency and sequence four times as much, the fold coverage will be 4-times higher but the mapping efficiency is unchanged...

[Hedi86]

thanks alot for your explanation, now became more clear