search

arefeen / TAPAS

17 stars 4 forks source link

TAPAS on GTF from Gencode v28 #6

Open JamalEH opened 4 years ago

JamalEH commented 4 years ago

Dear Team,

Thank you for the great tool TAPAS.

I have aligned my RNA-Seq reads against human genome using (GRCh38) genome version and gene annotation model from Gencode v27.

Can I provide the same GTF file to TAPAS to perfrom APA detection and differential expression analysis? Or should I first preprocess my GTF annotation file before running TAPAS?

Thank you so much! Kind regards, Jamal.

arefeen commented 4 years ago

Hi Jamal,

I am glad to known that you like our tool TAPAS. And, I am extremely sorry for the late reply.

In order to use your GTF file, you have to convert it into refFlat format. Please see the refFlat file in the "sample" directory for reference. Use the converted refFlat file to run the tool.

Thanks, Ashraful

JamalEH commented 4 years ago

Hi Arefeen, Thank you for your kind reply!

I successfully ran TAPAS on my dataset following your instructions. In my results output file, I observed no significance evaluation of the differential APA usage "shortening/lengthening" between my conditions. Is there any way to compute a significance "p-value" for the observed differences in APA length between the 2 conditions, since we have 3 samples per condition?

I also observed a lot of rows with NA values.

Could you please comment on this? Thank you! Kind regards, Jamal.

JamalEH commented 4 years ago

I may be missing some steps in my analysis.

I will restart a new fresh session and get back to you very soon. Can I produce my reflet file using this command? gtfToGenePred -genePredExt Gencode.v28.annotation.gtf Gencode_reflat.genePred

Kind regards, Jamal.

arefeen commented 4 years ago

Hi Jamal,

Try to match the output from gtfToGenePred with the sample refFlat file.

Thanks, Ashraful

Brycealong commented 4 months ago

Hi Arefeen,

Thank you for your excellent tool.

I noticed the program filters out non-coding genes by keep gene names like 'NM_XX'. But at the same time this also eliminates non-'NM'/'NR' prefixed entries from my customed refFlat file converted from a GTF file. I'm currently sticking with the official UCSC refFlat file, but it would be great if future versions could support a broader range of annotation formats.

Appreciate your work and looking forward to updates.

Best, Bryce