EvidenceModeler output

[Marco Sollitto]

Hi, 

I have a problem related to evidencemodeler output. I ran it on 3 different gff files (results of braker, genomethreader and stringtie). 

For checking, I ran a BUSCO analysis on the merging fasta files deriving from the previous gff3 files. I reached ~98% of gene presence.  

When I ran BUSCO, on the output of EvidenceModeler, I reach only the ~86%. How is it feasible? Where is the problem? 

How I can ameliorate its result? 

Thanks,

Marco

[戴睿紘]

Hi there,

I also have this problem.
My braker's BUSCO reached 92% and Maker's BUSCO is 80%. But the BUSCO of EVM result only got 70%.

My weight file set equal to these predictions.

Thanks for this question and people who helps.

Best,

Jui-Hung

marco.s...@gmail.com 在 2022年3月3日 星期四下午10:55:33 [UTC+8] 的信中寫道：