Finding sequence family specific primers using DECIPHER
1
0
Entering edit mode
9 weeks ago
Marcel • 0

Hello,

I am trying to find PCR primers that are specific to a family of repetitive elements in the human genome.

For this, I found the R package DECIPHER, which seemingly can do exactly this using its TileSeqs and DesignPrimers functions as described in one of their vignettes.

After getting DECIPHER to work and sucessfully designing specific primers for a test dataset, I aligned all the sequences for the repeats of my family of interest and also all sequences of closely related repeat families. I then told the program the family for each repeat sequence such that it can later discriminate between them (as described in 3.4 Defining Groups in the above vignette).

I then ran the program with:

tiles <- TileSeqs(dbConn, add2tbl="Tiles", minCoverage=1)
primers <- DesignPrimers(tiles, identifier="repeat_family_of_interest", minCoverage=1, minGroupCoverage=1)


However, the output was empty. I guess, this means, it's not possible to find a primer that will amplify all repeats of my family of interest, but not a single one of the other families. This is quite likely as there are thousands of repetitive elements that are highly polymorphic in my dataset and as I mentioned above, it had previously worked for a test dataset.

Now, I am also fine with a primer that will only amplify a fraction of my family of interest, but none of the other families. I am struggling in achieving this. I guess, this means, I need to set the minCoverage and/or minGroupCoverage parameters somehow accordingly. But even after reading the manual, I don't seem to understand what these parameters do exactly and how I'd need to set them to achieve what I want.

It would be great if someone could help.

DECIPHER PCR primer • 216 views
2
Entering edit mode
9 weeks ago
Erik Wright ▴ 410

Thanks for your interest in DECIPHER's DesignPrimers function.

You should be able to use it for this task as long as the target group is not too polymorphic. There might not be any viable primers and that is why the output is empty. You can look at the tiles assigned to each target group to see whether it might be feasible.

The argument minCoverage simply determines how much of the target group must be amplified by the primers. The problem is that sometimes sequences are not full length, so they might not have any information in the targeted region of the alignment. The minGroupCoverage argument controls the fraction of the sequences in the target group that must span this region of the alignment.

I hope that helps.

0
Entering edit mode

Thanks for clarification of what these parameters do. I will try to improve my alignment and then gradually lower the minCoverage paramerer and see whether I can find primers.