Incomplete TE identification by repeatmasker

Question

[jaqx008]

Hello,

I am trying to obtain all TEs from a genome using RepeatMasker. However, the result only returned simple repeats and low complexity TEs (1.71% total). All the other TE classes returned zero values. Am I doing something wrong? How can I recover all the other TE classes? See my command below:

RepeatMasker -pa 6 -species Branchiostoma belcheri genomic.fna > Masked_Bbe

Thanks

[samuel.a.odonnell]

This will most likely be due to a library issue, where the TE library repeatmasker is searching with is basically non-existant. I had the same issue which I think was due to my conda installation...are you perhaps in the same boat?

I don't know how commonly studied your species is but your alternative is generating a denovo repeat library using RepeatModeler (as perhaps the generic repeatmasker repeat libraries would poorly represent your species).

[jaqx008]

That makes sense. I did try Repeatmasker but the sequences returned are not grouped into TE classes or types. just fasta sequences representing the masked TEs. Is this typical with RepeatModeler?

[samuel.a.odonnell]

I am not sure I understand. You are saying you tried de novo TE identification with RepeatModeler?

[jaqx008]

Yes I did. But the output was a bunch of repeat sequences in fasta format. The headers didnt say what type of TE class was discovered. This is the command I used after I created a db from the genome.

RepeatModeler -database Bbe engine ncbi -pa 8

Do you have a better command I can try?

[samuel.a.odonnell]

So all the repeats identified by RepeatModeler are labelled "unknown"? Otherwise the fasta header for each sequence would have an indication of the repeat type.

Unfortunately, that is exactly how I would run RepeatModeler also and the fasta file should be called 'Bbe-families'. After running something like BuildDatabase -name Bbe -engine ncbi Bbe.assembly.fa of course