How to treat with high mismatch reads in whole genome studies?

Question

[Whoknows]

Hi all friends

I have whole genome project, during this alignment with human genome I faced with a big challenge, how to treat with un-mapped reads which have high mismatches for example 20-30 mismatches in a read. Because, usually aligners discard them and save them in an un-mapped reads file.

But my questions is, Is there any way to do more work on them or I have to ignore them?

Many thanks

[Pierre Lindenbaum]

you already opened many question: having up-regulated and down-regulated isoforms in Trinity output , Tophat2 shows error during running , Guidance for a personal medicine data center , Error in samtools uniquely mapped reads file, Finding isoform expression in RNA-Seq ... but you did not validate them.