[question] reads detected sequences out of sequencing primer in the other direction

[SF_mallish]

Hi,

I've got a pair-end illumina sequencing data (100bp each end) for Ago iClip in mouse ES cell.
majority of read pairs show that the inserted fragments are very short. As shown in the following example:

@DBRHHJN1:278:C11RFACXX:3:1101:1443:1167 1:N:0:
ATAGGTATGCGCCACTGAGATCGGAAGAGCGGTTCAGCAGGAATGCCGAGACCGATCTCGTATGCCGTCTTCTGCTTGAAAAAAAAAAAAAAAAGCACAC
@DBRHHJN1:278:C11RFACXX:3:1101:1443:1167 2:N:0:
CAGTGGCGCATACCTATAGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGTAGATCTCGGTGGTCGCCGTATCATTAAAAAAAAAAAAAAACACAAGAGAG

Above is one read pair. The underlined part of each read can be aligned with illumina PCR primers in the other end. And the 17bp part of each read is reverse complement to each other. This means that the insert fragment is only 17bp and the 100bp read go through the PCR primer (and sequencing primer) in the other end.

Could any one explain me what is the "AAAAAAAAAAAAAAAAGCACAC" and "AAAAAAAAAAAAAAACACAAGAGAG" after the underlined sequences? How can I interpret them?
The fastq score is low for these regions.

Here is the same reads with fastq score:

@DBRHHJN1:278:C11RFACXX:3:1101:1443:1167 1:N:0:
ATAGGTATGCGCCACTGAGATCGGAAGAGCGGTTCAGCAGGAATGCCGAGACCGATCTCGTATGCCGTCTTCTGCTTGAAAAAAAAAAAAAAAAGCACAC
+
@C@FFDDFHHHHHJJJJJIIJJJJJJJGHJJJJGIJJJJJJIJIJIIJIGHGGFC>ADBD@CDDDDDDDDDDDCCDCBDDDBDDDDB############
@DBRHHJN1:278:C11RFACXX:3:1101:1443:1167 2:N:0:
CAGTGGCGCATACCTATAGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGTAGATCTCGGTGGTCGCCGTATCATTAAAAAAAAAAAAAAACACAAGAGAG
+
CCCFFFFFHHHHHJJIJJJJJJJJJIJIHGIJJJJJIDIIIJJJJJIHIJCHFHHHHHFFFDDEDDDDDDDDEEEDDCCCBDB#################

Also, I am curious what will be the reason for these short insert fragments during sequencing. After library construction, the results of bioanalyzer show that the average length of insert fragment is ~100bp.

[kmcarr]

Could any one explain me what is the "AAAAAAAAAAAAAAAAGCACAC" and "AAAAAAAAAAAAAAACACAAGAGAG" after the underlined sequences?

After sequencing through the Illumina adapter the polymerase has literally reached the end of the line. There are physically no more bases being incorporated yet the software is "forced" to make a basecall. It's the equivalent of a wild guess.

How can I interpret them?

You can't, they're meaningless.

Also, I am curious what will be the reason for these short insert fragments during sequencing. After library construction, the results of bioanalyzer show that the average length of insert fragment is ~100bp.

It is very easy for peak of small/no insert fragments to hide in a BioAnalyzer trace. For extremely small library molecules like this a small physical amount of DNA can translate into a large molar amount.

[SF_mallish]

Thanks kmcarr for your quick reply!

For the last point you mentioned,

For extremely small library molecules like this a small physical amount of DNA can translate into a large molar amount

Could you explain more detail about it, or do you have any reference for that?

The experiment was done by my lab mate. Actually, he measured the length of the total fragment, which contain ~70bp adaptors (like sequences for PCR primers) flanking the inserted fragment in both end. The peak of bioanalyzer result is around 240bp, so we interpret the average inserted fragment size is ~100bp. Will the results of this protocol still be affected by the reason you mentioned?

Thanks!