Trouble: getting reads with Reverse primer

[bio_informatics]

Hi SeqA team,

I've hit another road block.
Data information: 16s V3-V4 region, demultiplexed. Illumina, Paired end. PhiX 5%

I've forward, and reverse primer information.
F: CCTACGGGDGGCWGCA
R: GGACTACHVGGGTMTCTAATC

When I try to grep sequences with Forward primer in R1, I get reads from 1k-12k. With all degenerate primer combination.

However, I get less than 10 sequences with reverse primer, degenerate combination on R2.
I tried with ^ (starts), ($)ends with, nothing seems to give me reads for reverse primer.
Tried with reverse complement of reverse primer, no results.

1)
Wouldn't R2 sequences have reverse primers?

2)
I went ahead with forward, and reverse primer information to assemble the reads, with Mothur. I didn't get any output.
When I removed reverse primer information, I managed to get an output of decent file size.


The 16S sequencing was done as the protocol

[GenoMax]

Have you tried to find the reverse complement of R (i.e. GATTAGAKACCCBDGTAGTCC) with either reads?

[bio_informatics]

Hi genomax,
Thanks for your reply.

I tried with degenerate primers of reverse complement (GGACTACHVGGGTMTCTAATC).
No outputs. 0 .

I double checked with the one you shared. No outputs. 0.

I double checked with sequencing center, they said they have used R as reverse primers. And no other processing has been done apart from index removal (during de-multiplex).

Weird.

[Brian Bushnell]

By the way, BBDuk processes degenerate bases (with the copyundefined flag), and looks for both the forward and reverse-complement... you might find that to be a more robust alternative to grep.

bbduk.sh in=reads.fq out=matching.fq literal=CCTACGGGDGGCWGCA k=16 mm=f copyundefined=t

To get reads containing both the F and R sequences would require 2 sequential passes.

[GenoMax]

@Brian @bio_informatics could also use "kmercountexact.sh" to see what oligo is overrepresented? In case the sequence of R is incorrect?

[bio_informatics]

Hi Brian,
Thanks for adding information about bbduk. I shall try with that, too.

But the question haunts. Why there are no reads with reverse primer: either way - forward or reverse complement?

[bio_informatics]

Hi genomax,

Thanks for the idea.
I pulled out fastqc, and in that looked for overrepresented sequences.
For reverse read.

Of many sequences, I've:
Over:
GACTACTGGGGTATCTAATCCTGTTTGATCCCCACGCTTTCGCACATCAG

Rev Primer:
GGACTACHVGGGTMTCTAATC

Clearly this is unmatched until the length of Reverse primer. (even leaving degenerate bases)

@Brian @bio_informatics could also use "kmercountexact.sh" to see what oligo is overrepresented? In case the sequence of R is incorrect?

[Brian Bushnell]

@Brian @bio_informatics could also use "kmercountexact.sh" to see what oligo is overrepresented? In case the sequence of R is incorrect?

That's true, but since many parts of the V3-V4 region are so highly conserved... and, well, with amplicon sequencing in general, it might be hard to get any useful signal unless you restrict the search to only the end of the read, by trimming everything other than the first 22bp before the analysis, then reverse-complement the reads and repeat to get the other end.

You should actually be able to just do this visually - the primer sequence in question should be the first or last 22bp. What do you see there?

[bio_informatics]

You should actually be able to just do this visually - the primer sequence in question should be the first or last 21bp. What do you see there?

Thanks for your reply.
I pulled out few sequence, trim until 21 bp. Below are 3 lines after trimming 21:

GACTACTCGGGTCTCTAATCC
GACTACTTGGGTATCTAATCC
GACTACAAGGGTCTCTAATCC

GGACTACHVGGGTMTCTAATC

To which my reverse primer is terrible in matching with them.

[GenoMax]

Hopefully the right region has been amplified in that dataset.

[bio_informatics]

Hopefully the right region has been amplified in that dataset.

Amen!, I too hope the same.
Primer is clearly different then the ones I'm able to see.

[GenoMax]

Someone must have used a wrong primer. Not much you can do but report to experimental folks.

[Brian Bushnell]

Actually, that's a perfect match:

 GACTACAAGGGTCTCTAATCC
[B]GGACTAC[U]HV[/U]GGGT[U]M[/U]TCTAATC[/B]

They only differ at the degenerate symbols, and the degenerate symbols are:

H: A or C or T
V: A or C or G
M: A or C

...which include the bases in question. For some reason it's offset by one base though. But that's probably not an issue with the primer, just where reading starts.

[bio_informatics]

Actually, that's a perfect match:

 GACTACAAGGGTCTCTAATCC
[B]GGACTAC[U]HV[/U]GGGT[U]M[/U]TCTAATC[/B]

They only differ at the degenerate symbols, and the degenerate symbols are:

H: A or C or T
V: A or C or G
M: A or C

...which include the bases in question. For some reason it's offset by one base though. But that's probably not an issue with the primer, just where reading starts.

Thanks. This is what I have used for all the data received. All files have trimmed primer.