Can the "QNAME" information of a SAM file be used for estimate uniquely mapped reads?

[znasim09]

Hello everyone,

First of all, I know there has been a lot of discussions related to uniquely mapped reads estimation (sorry for bringing it up again).

My question is that QNAME, the first column of SAM file, i.e.

Reads/segments having identical QNAME are regarded to come from the same template

Can this information be used for estimating the uniquely mapped reads?

For example, for a SAM file from a paired-end dataset, uniquely mapped reads can be differentiated as the ones that occurs "twice" (two same QNAMEs). Similarly, if there are more than two same QNAMEs, they can be "multi-mapped reads". And if there is only one QNAME, it can be classified as "broken read".
I checked it with the output statistics from STAR aligner and it perfectly matches with it (% of uniquely and multi mapped reads), however, i am not sure about other aligners and don't know if this way is acceptable.

Can anyone please let me know if this is an acceptable way or not? And/or suggest a better way PS I am just an ordinary biologist without much bioinformatics background.

Thanks