No CDS found (proteinToGenome)

[Zupanova]

Hello,
I am trying to map protein coordinates to genome coordinates for Arabidopsis Thaliana. I tried both ensDbFromGtf() and ensDbFromGff() to create the ensDB object. As protein ID I used the given transcript ids "tx_id".
Although the ID and CDS coordinates are definitely contained in the file (checked manually) I keep getting this message:
Fetching CDS for 1 proteins ... 0 found
Checking CDS and protein sequence lengths ... 0/0 OK
Warning message:
In proteinToGenome(x, EDB_Ath, id = "name", idType = "tx_id") :
No CDS found for: AT1G01470.1

What could be the cause?

P.S I also tried ensDbFromAH() but there was no GTF file give for A.Thaliana among the 7 found

Greetings

[jorainer]

Hi, the problem is that the protein annotations needed for proteinToGenome can not be imported from Gtf files. These are only available if the EnsDb is created directly with the Ensembl Perl API. If you tell me the Ensembl version (or better the Ensemblgenome release version, e.g. 47) you would need I will build the EnsDb for you.

[Zupanova]

Hi, thanks for the reply. It would be very nice if You could do that for me. I am working with Arabidopsis Thaliana TAIR 10 sequences, so release version 47 on Ensembl.
I still have one question: for what purpose can the functions ensDbFromGtf() and ensDbFromGff() be used?
Greetings

[jorainer]

You can download the EnsDb SQLite file from here. You can then simply provide the file name (inclusive path) of this file in the EnsDb constructor to load the resource (i.e. edb <- EnsDb(<sqlite file>)) and use it right away.

[Zupanova]

Ok. Thank You very much for your help.

[jorainer]

Closing this issue now, feel free to re-open if needed.