any scripts to count percentage of human reads from kneaddata's output?

[Ming Liao]

Hello, 

I have got the clean data for each sample after running kneadata. I think the next step is to summarize how many human reads are contained in the original raw data. Is there any script for this task? Thanks.

Best,

Ming 

[Andy Shi]

Hi Ming,

If you look at the printed output of kneaddata or the log file, you can see some output containing the original read count, and the read count after merging from multiple databases. For the demo, the initial read count was 1075, and at the end there were 1065 reads. This can tell you how many reads were removed.

Best,

Andy

[Husen Zhang]

Adding to Andy's email, I use "grep alignment *kneaddata.log" to get the numbers. Then yes you will still need to do simple math to get %.