MiSEQ low diversity full Overlapping Paired end reverse read problems

[fgilpara]

Hello, im new here and i have a question about a sequencing problem we were getting.

For a project, we use Miseq and nano kit for doing 2*151 paired sequencing.

We sequence 1 or 2 amplicons of known coding with 1 snp of interest.

in each run we sequence 3 o 4 biological samples as duplicates or triplicates for a total 13 barcodes of 6 bp).

The design is made this way to get a 100% overlap between R1 and R2 cause we need accuracy calling the SNP (the original sample has a very low dna suppossed to carry the snp and high quantity of DNA know to be WT).

Important to say that basically on the PCR to add the adapters and barcodes the amplified sequence is only 2 bp (including the snp position). And we use 20% phix to compensate the low variability.

The problem:
For one of the amplicons (runing alone or with the other amplicon) everything is okay- A little low R2 quality but fine. also, cause the insert size is lower than 151 we get some low quality poly A and the end of the reads but thats not a problem.

For the other amplicon (another gene) the forward read is OK but the reverse read is a nightmare. We don't know whats happening.

Attached is the FastQC report for 1 amplicon_good and amplicon_bad R1 and R2, from the same run.

Thankyou

[nucacidhunter]

I do not know your library prep details but it seems that bad amplicon R2 has not primed well possibly due to some base mismatch in adapter sequences.

[fgilpara]

1 PCR to add adapters and 1 index. Forward common, reverse different index.


AATGATACGGCGACCACCGA
Adapter P5

GATCTACACTCTTTCCCTACACGACGCTCTTCCGATCTTAGCTAGC
RD1

TTGGGCTCCTGTCTTACAGGC
gene specific primer 5’

CCTGCCTCCGGGCTCACCTCGCTGTGACCT Amplified sequence

GAAGGAGAATCTGCTGAAGGATAACT
gene specific primer 5’

GATCGGAAGAGCACACGTCTGAACTCCAGTCAC
Primer RD2

XXXXXX
Index

ATCTCGTATGCCGTCTTCTGCTTG
Adapter P7

[nucacidhunter]

I wonder if you could attach the sequence of the final amplicon with adapters that goes into sequencer. I assume you are using standard Illumina sequencing primers included in sequencing kits.

[fgilpara]

Yes we use illumina universal primers.
The data in the last post is amplicón one 5'-3> as it enters the cartridge.

Yes we use illumina universal primers.
The data in the last post is amplicón one 5'-3> as it enters the cartridge. (with its reverse complement too.

AATGATACGGCGACCACCGA
GATCTACACTCTTTCCCTACACGACGCTCTTCCGATCTTAGC

TTGGGCTCCTGTCTTACAGGCCCTGCCTCCGGGCTCACCTCGCTGTGACCTGAAGGAGAATCTGCTGAAGGATAACT

GATCGGAAGAGCACACGTCTGAACTCCAGTCAC
GCCAAT
ATCTCGTATGCCGTCTTCTGCTTG

in order: P5 adapter to flowcell, RD1 adapter (here hybridizes read primer 1), GSP5' - interest portion -GSP3' (gene speceofic primers), RD2 adapter (here hibridize first the index primer and later the sequencing primer 2), index, Adapter p7 to flowcell.

The reverse complement is ommited for clarity.

The question is our adapter sequences (P5,P7, RD1, RD2) are the same for the two amplicons, and only is failing the reverse read of one of them.

[fgilpara]

As for the libprep, we buy the primers taht consist on 1 forward primer which includes the 5'-P5 adapter + RD1 adapter + GSP3'

and some reverse primers wich consist in: 5'-P7 adaptor + index + RD2 + GSP 3' varying only in the 6bp index.

This is for the 2 amplicons (1 of gene A exon X and another for the Gene B exon Y)

After amplification, agarose gel to verify amplification, purification with ampureXP, quantification and standarization, pool mixing denaturation with NaOH. All goes to 16 well of the cartridge .

[fgilpara]

For clarity i made a paint scheme.

[nucacidhunter]

It seems that your P7 adapter is missing an A at the start. The A is not part of adapter but it is added during A tailing in shotgun library prep. Since you are adding adapters as part of your GSP you should add the A, otherwise the fragments will be primed by truncated sequencing primers from multiple positions in a cluster which will result in low quality sequences.

I guess the other amplicon that sequences well accidentally has an A in that position which is part of your primer.

AATGATACGGCGACCACCGA GATCTACACTCTTTCCCTACACGACGCTCTTCCGATCTTAGCTTGGGCTCCTGTCTTACAGGCCCTGCCTCCGGGCTCACCTCGCTGTGACCTGAAGGAGAATCTGCTGAAGGATAACT


AGATCGGAAGAGCACACGTCTGAACTCCAGTCAC GCCAAT ATCT CGTATGCCGTCTTCTGCTTG

[fgilpara]

Som additional Info: Fastq from bad_amplicon_forward_replica1 and bad_amplicon_reverse_replica_1. its a .zip

[fgilpara]

Here is the other amplicon for clarity. That one is getting sequenced perfectly in forward and in reverse.

AATGATACGGCGACCACCGA P5
GATCTACACTCTTTCCCTACACGACGCTCTTCCGATCTTAGC RD1
GTAAATGGACTTCCTTAAACTTTAACCGA GSP 5'
AC Amplified seq
GCTGAGACTTCTGATGAGTCAGTATGG GSP 3'
GATCGGAAGAGCACACGTCTGAACTCCAGTCAC RD2
AGTTCC Índex (XXXXXX)
ATCTCGTATGCCGTCTTCTGCTTG P7

[fgilpara]

It seems that your P7 adapter is missing an A at the start. The A is not part of adapter but it is added during A tailing in shotgun library prep. Since you are adding adapters as part of your GSP you should add the A, otherwise the fragments will be primed by truncated sequencing primers from multiple positions in a cluster which will result in low quality sequences.

I guess the other amplicon that sequences well accidentally has an A in that position which is part of your primer.

AATGATACGGCGACCACCGA GATCTACACTCTTTCCCTACACGACGCTCTTCCGATCTTAGCTTGGGCTCCTGTCTTACAGGCCCTGCCTCCGGGCTCACCTCGCTGTGACCTGAAGGAGAATCTGCTGAAGGATAACT


AGATCGGAAGAGCACACGTCTGAACTCCAGTCAC GCCAAT ATCT CGTATGCCGTCTTCTGCTTG

I know my adapters are from an older version and i see the new ones have and aditional A like you say but i don't fully understand the problem you are trying to point out. Could you please elaborate or make a small pic? i thank you so much for helping me.

[nucacidhunter]

I assume the sequences that you have provided represent top strand of amplicon if put together continuously as following.

AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCTTAGCTTGGGCTCCTGTCTTACAGGCCCTGCCTCCGGGCTCACCTCGCTGTGACCTGAAGGAGAATCTGCTGAAGGATAACTGATCGGAAGAGCACACGTCTGAACTCCAGTCACGCCAATATCTCGTATGCCGTCTTCTGCTTG

If it is sequenced to the end should give following R1 sequence:

TAGCTTGGGCTCCTGTCTTACAGGCCCTGCCTCCGGGCTCACCTCGCTGTGACCTGAAGGAGAATCTGCTGAAGGATAACTGATCGGAAGAGCACACGTCTGAACTCCAGTCACGCCAATATCTCGTATGCCGTCTTCTGCTTG

It seems that sequences that you have provided in posts above is not representing your actual amplicon sequence. A typical sequence from your sequence file is as following (I have trimmed sequences after end of adapter which includes polyA and some other background signal):

TAGCTTGGGCTCCTGTCTTACAGGCCCTGCCTCCGGGCTCACCTCGCTGTGACGCTCGGACGAGCACACGTCTGAACTCCAGTCACATGCGCATCTCGTATGCCGTCTTCTGCTTG

For your information, the A is the base complementary to the most 3’ base of sequencing primer. Please look at page 2 of attached HT adapters where non-adapter A added during A tailing of shotgun library prep is shown. These adapters are only different from LT adapters in index and have few extra bases that are not part of R1 or R2 sequencing primers. Sequencing primer have a T at 3’ end which pairs with that A and if A is absent then it cannot be extended during sequencing. R2 sequences then will be read by truncated R2 sequencing primer resulting in erroneous and low quality bases. This is R2 sequencing primer: 5’ GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT and note the 3’ T.

[fgilpara]

You are correct, the fastqc i provided in .gz file are from a new run using new primers with shorter sequence specific region cause we thought the problem could be some part of the gene sequence that caused a secondary structure to be formed and that affected the reverse read.

Thank you for your time and help, i will update further.

[fgilpara]

nucacidhunter, thankyou, you were right, the sequence lacked that nucleotide.
Im trying to send you a private message but its not working:

Thank you, you were right. Even illumina support was telling us that the original sequence was ok but some experiments worked better with that nucleotide addition, but that they cant help us further cause its a custom primers experiment. Your solution worked, we adquired primers with the lacking nucleotide and it worked like charm. We suspect that the mismatch in the another amplicon was not sufficient to make it fail but changed that primers too. Thanks!!!!!