how can interpret these biologically weird results?

0

Entering edit mode

6.4 years ago

Mozart ▴ 330

Hello there, as final results of my pipeline (sleuth) I obtained a redundancy in terms of different transcript related to the same gene, up and down regulated at the same time.

Any hints on how to interpret these results?

RNA-Seq • 1.9k views

ADD COMMENT • link updated 3.2 years ago by Biostar 20 • written 6.4 years ago by Mozart ▴ 330

0

Entering edit mode

                 target_id           ens_gene ext_gene
38604 ENSMUST00000194462.5 ENSMUSG00000027737  Slc7a11
38492 ENSMUST00000193838.1 ENSMUSG00000027737  Slc7a11
3477  ENSMUST00000029297.5 ENSMUSG00000027737  Slc7a11
25123 ENSMUST00000142932.2 ENSMUSG00000027737  Slc7a11

ADD REPLY • link 6.1 years ago by Mozart ▴ 330

4

Entering edit mode

Thanks for reposting here. It's just what I had expected was happening here.

Look at the var_obs column. The variance for ENSMUST00000194462.5 is completely inflated, ~8-9 times greater than the others. This will result in unreliable fold-changes and P values.

rss and tech_var neither look so hot.

ATPoint, what do you think?

ADD REPLY • link 6.4 years ago by Kevin Blighe 87k

0

Entering edit mode

Sorry but if I am not mistaking this means that when you filter your result the most significant (statistically and by fold change) is the only one I should consider?

ADD REPLY • link 6.4 years ago by Mozart ▴ 330

1

Entering edit mode

Yes of course. The general rule is:

absolute log2 fold change >2

and

FDR Q value (adjusted P value) ≤0.05

The high variance of the transcript resulted in it not passing FDR adjustment, highlight just how important these adjustments are.

ADD REPLY • link 6.4 years ago by Kevin Blighe 87k

0

Entering edit mode

Thanks a lot...and sorry if sometimes I am not that precise.

ADD REPLY • link 6.4 years ago by Mozart ▴ 330

1

Entering edit mode

Absolutely no problem my friend. If you are indeed the real Mozart, then keep up the great music!