Status of maker annotation

Question

[Prakki Rama]

Hi all,

May I know if there is any tweak/trick to know the status of genome annotation using maker? Annotation of 1 gb genome using maker has been running for past two months on 24 cores. May I know if there is a way to speed up the annotation for larger genomes? Please share your comments and valuable suggestions.

Answer 1

[Juke34]

You can have an idea of the status of the annotation by comparing the number of finished contig in the log file with the number of contig of your assembly.

The biggest part of the computational time is used by the alignment of the evidences in fasta format. So, to speed up the annotation you may reduce the data amont used to feed MAKER. What you can do also, is to perform the splice aware alignment outside MAKER with more performant tool (e.g GMAP) and feed up MAKER with the gff produced.

[Prakki Rama]

Hi. Thank you for suggestions. They were really helpful. Some of the genome scaffolds were showing FAILED status. Should these be extracted and run maker again seperately and finally merge the gff files at the end. Thanks in advance for your suggestions.

[Juke34]

By default MAKER try to annotate a chunck two times. After these two tries it skip it. No need to extract the failed sequences. When the MAKER run is finished you can just relaunch it whith the parameter that allows MAKER to retry to annotate the failed sequences more than 2 times. Either whith the option "-t 6" (6 means try six times), or modifying the retry option whitin the maker_opts file (around the end of the file).

[Prakki Rama]

Thank you. I will try that!

edit: it was useful thank you!

[Prakki Rama]

I could check the number of finished chromosomes/contigs from the data store folder using the following command:

ls -ltrh */*/*/*.fasta | awk '{print $NF}'| awk -F'/' '{print $3}' | sort -u | wc -l