Marc, just trying to clarify something I have read in on the posts.
Can use Mercator to generate our own mapping files from a de novo transcriptome assembly ? The species in question does not have a genome sequence yet.


Thank you

Jose

Yes, exactly that is what mercator was made for - if your assembly is too large
for submission via the web interface you can send it to me via:
https://cryptshare.mpg.de/index.php?lg=en

cheers,
Marc

Thank you Marc.
We will try the online submission but our assemblies have 100,000-400,000 contigs

Anything above 30,000 will be rejected by the web interface, i fear, so you can save
yourself the effort. 100,000 to 400,000 is very large... are these meta-transcriptomes
or are you expecting a single organism to actually have this number of genuine
transcripts?

Maybe it is worth considering prefiltering the data and removing all sequences that are
shorter than, say, 100nt? How many would be removed? You could also consider
removing all sequences that are just singletons (not supported by more than one
read). These steps might help purge a lot of non-significant data from primary trancriptome
assemblies.

best greetings,
Marc

Marc, these are single organism assemblies. I know one organism will not have that many genuine transcripts.
In all cases the organisms are polyploid,.. All of them are grasses. In all cases are contigs, not singletons, over 300 bp. In the case of 400,000 contigs the assembly is the output resulting from over a billion read assembly, from reads from several individuals, using trinity, so the output contains different isoforms (these could be paralogs, alleles, etc..). He have done the GO annotation, so may be just keeping those with a GO annotation?

Hi Jose,

sorry if my previous mail sounded harsh - i just asked to learn more
about your dataset. No - i would not suggest to leave out those without GO
annotation. We can run the whole dataset through mercator - you'll just have to be
a bit more patient since the processing will take a bit longer than for a "standard-size"
transcriptome. Just send the data via the link i posted above and i will submit it
to the pipeline.

cheers,
Marc

Don't worry Marc.
I will send them,.. One at a time thru the link you provide me. I will have one my students doing the first one ,.. And smaller...

Thank you

Jose