Genome size and corrected genome size

[priya]

Hi,
I came across this term "corrected genome size" while reading one paper. Is there any difference between genome size and corrected genome size. If yes, then what needs to be corrected for genome size.

[amitm]

hi priya,
You haven't given any specific detail. Was the term used in context to NGS analysis? Was it referring to a reference genome or a de novo assembled one?

[priya]

hi priya,
You haven't given any specific detail. Was the term used in context to NGS analysis? Was it referring to a reference genome or a de novo assembled one?

I came across this term in paper describing normalization of chip-seq reads .
For your better understanding , I attached screenshoot of lines from the paper .

[amitm]

hi,
it seems that corrected genome size is the area of the genome covered by all the ChIP-seq reads in that sample.
So, if sample A has 20M reads and they cover 2Gb of hg19, then corrected genome size is 2Gb.

[priya]

hi,
it seems that corrected genome size is the area of the genome covered by all the ChIP-seq reads in that sample.
So, if sample A has 20M reads and they cover 2Gb of hg19, then corrected genome size is 2Gb.

Hi amitm,
Thank you for your reply!

Can you please clarify me how to calculate the genome coverage from sequencing experiment.
For sample read coverage, i can easily check the alignment logs (for eg: Bowtie log files), which gives me clearly stat of number of reads mapped per sample.

[amitm]

hi,
Once you have done the mapping of reads, use the BAM file obtained to create a BED file.
Use bedtools -

http://bedtools.readthedocs.org/en/latest/content/tools/bamtobed.html

Then, the coordinates returned would be overlapping. You need to merge them to create "unique" regions/ coordinate intervals.
Use -

http://bedtools.readthedocs.org/en/latest/content/tools/merge.html

Once there, add up the lengths of all intervals and thats the portion of the genome covered, i.e. corrected genome size

[priya]

hi,
Once you have done the mapping of reads, use the BAM file obtained to create a BED file.
Use bedtools -

http://bedtools.readthedocs.org/en/latest/content/tools/bamtobed.html

Then, the coordinates returned would be overlapping. You need to merge them to create "unique" regions/ coordinate intervals.
Use -

http://bedtools.readthedocs.org/en/latest/content/tools/merge.html

Once there, add up the lengths of all intervals and thats the portion of the genome covered, i.e. corrected genome size

Hi amitm,
Thanks alot for your clear explaination. I will try it out

[AlexReynolds]

You can use BEDOPS bam2bed to convert from BAM to BED, pipe to bedops to merge overlapping elements, and pipe to bedmap to generate a list of lengths per merged element to sum with awk:

$ bam2bed < foo.bam | bedops --merge - | bedmap --echo-overlap-size - | awk '{s += $1;} END {print s;}' > answer.txt

In this case, bedmap is mapping merged elements against themselves. Merged elements coming out of bedops are guaranteed to be disjoint, so --echo-overlap-size is guaranteed to report the unique length of each merged element.