[maker-devel] Format error detected with maker2eval_gtf

[Zoe Clarke]

Hello!

I am currently working with an annotation from Maker that has just finished. Unfortunately I think there are some errors in my annotation, as a lot of my reads from a test RNA-seq alignment got discarded from my analysis. The analysis tool suggested my annotation might have overlapping reads.

I am trying to figure out what went wrong with my annotation, and the closest hint I have are error messages that I receive from running maker2eval_gtf on my output merged gff file. It appears every line of my gff is producing the following errors:

Use of uninitialized value $att in pattern match (m//) at maker2eval_gtf line 244, <IN> line 200199481.

Use of uninitialized value $att in pattern match (m//) at maker2eval_gtf line 245, <IN> line 200199481.

Use of uninitialized value $att in pattern match (m//) at maker2eval_gtf line 246, <IN> line 200199481.

Use of uninitialized value in string eq at maker2eval_gtf line 35, <IN> line 200199481.

Use of uninitialized value in string eq at maker2eval_gtf line 37, <IN> line 200199481.

Use of uninitialized value in string eq at maker2eval_gtf line 38, <IN> line 200199481.

Use of uninitialized value in string eq at maker2eval_gtf line 39, <IN> line 200199481.

Use of uninitialized value in string eq at maker2eval_gtf line 40, <IN> line 200199481.

Use of uninitialized value in string eq at maker2eval_gtf line 41, <IN> line 200199481.

Do you know what issue this might indicate for my gff? At this point, I have done no filtering, just merged the files with gff3_merge.

Thank you so much, and please let me know if there is anything else I can provide to inform you more about my issue!

Zoe

______________________________________

Zoe Clarke

PhD candidate in Computational Biology at U of T

Lab profile: http://baderlab.org/Zoe%20Clarke

Personal website: https://zoe-clarke.weebly.com/

[Carson Holt]

Truncated line in the GFF3.  Network mounted file systems can sometimes return true on write operations that actually failed (should have killed the program, but NFS lies, and the program keeps running). Since MAKER does such high IO, this comes up from time to time. If you can identify the contig involved, you can either rerun it or delete just the gff3 files in the folder and let MAEKR rewrite them.

—Carson

_______________________________________________
maker-devel mailing list
maker...@yandell-lab.org
http://yandell-lab.org/mailman/listinfo/maker-devel_yandell-lab.org