Insert size calculation

[julia.fe...@gmail.com]

Dear Qualimap Team,

I have an issue with the mean insert size value calculation. The results are off - Qualimap reports a very large value which is definitely not correct - I calculated it independently with samtools and the histogram in the qualimap report also shows the correct distribution.

I am using the Qualimap version 2.2.1 and java 1.8.0_112. I also tried it with the old Qualimap version 2.0. It didn't report the correct value either. 

Example of results:

2.5. Insert size

Mean 2,288.59

Standard Deviation 236,689.68

P25/Median/P75 185 / 261 / 351

See attached file for histogram

Samtools:

Mean 558.3198

Median 261

Thank you very much in advance!

Best,

Julia 

[Konstantin Okonechnikov]

Dear Julia,

The variance between Qualimap and samtools results is most likely connected to the limits of samtools info for insert size. The default value is 8000.

Qualimap has not limit, therefore largest insert size is collected. In general, insert size data might include enough anomalous values from chimeras and other artifacts to make the mean value misleading regarding the real distribution. To see top of these values simply run the samtools command and sort insert sizes:

samtools view file.bam | cut -f 9 | sort -nr | head -n 10

However Qualimap also reports the P25/median/P75 values, which are computed from the same collected insert size array and give more precise estimation. The forming plot with border limits is also a good overview.

--

  Konstantin

--
You received this message because you are subscribed to the Google Groups "QualiMap" group.
To unsubscribe from this group and stop receiving emails from it, send an email to qualimap+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[julia.fe...@gmail.com]

Thanks for your answer!

To unsubscribe from this group and stop receiving emails from it, send an email to qualimap+u...@googlegroups.com.