I am a student in a Cancer lab. Working with sanger is new to me. While analyzing a report we found an insertion that has not been reported in any databases so far, we were working on checking if the insertion is pathogenic or not but when we reanalyzed the sanger annotation through variant reporter of the insertion it was giving it one position ahead and we analyzed that with the VEP as well that is were the confusion started.

The sequencer showed annotation of an insertion not between the bases but a step/position next to it:

c.5532-5533insGAA is the insertion position came after position 32 and before 33 but the sequencer annotates it like c.5533-5534insGAA.

We are working on human genome hg38.

Should I consider it a machine error or should I take that position for the result? Or is there a nomenclature error I am making here?

And the annotation is in this form [c.33-34insAUU]+[=] and I don't know what the +[=] represents here.

  • 2
    $\begingroup$ How do you know it should be c.32-33insAUU and not c.33-34insAUU? Could this be an issue of 0-based vs 1-based position counting? $\endgroup$
    – terdon
    Jan 5, 2023 at 11:25
  • $\begingroup$ I haven't heard of the 0-based and 1-based system. I read the linked post and it would make sense if that is the case. The insertion in that case would be 0-based, I assume. But it would be hard to say if that really is the case since I don't about this system. And it would possibly affect the result analysis, right? bcz the 1st annotation in my data tells its frameshift insertion while the 2nd annotation shows protein termination even when there is no stop codon there, according to Ensembl VEP results. (the annotation in the question are examples) $\endgroup$
    – user16732
    Jan 5, 2023 at 16:27
  • $\begingroup$ Please edit your question and add more detail. You said you were performing Sanger sequencing, but then mention VEP. VEP requires specific variants, so can only be used after sequencing. Please explain what you are really doing so we can understand. Basically, who is telling you c.32-33insAUU and who is telling you c.33-34insAUU? What input data is given in each case? $\endgroup$
    – terdon
    Jan 5, 2023 at 16:30
  • 1
    $\begingroup$ Try looking up the variant on a tool like varsome.com (disclaimer: I work for the company that builds that tool, but it's free) where you can see the underlying genome sequence, and can easily switch between hg19 and hg38. VarSome will also tell you if there are multiple, equivalent indels (various variants that all cause the same genotype, think of a deletion of A from AAA: no matter which A you delete, you end up with AA). $\endgroup$
    – terdon
    Jan 5, 2023 at 16:53
  • 1
    $\begingroup$ Thank you for your guidance, I'll redo my post once I gather more specific information. $\endgroup$
    – user16732
    Jan 5, 2023 at 16:58

1 Answer 1


Just to give my thoughts:

  • The missing information is the actual sequence
  1. Is this a conflict in the change in position of the insert?
  2. Could it simply be a consequence of down stream numbering, either a logical error (coding stuff), or lab artefact.

My bet is its point 2, but really we don't know except for the OP because we don't know the sequence.

Sources of error, Point 2

  • Sanger sequencing was traditionally known for "frame-shift" errors if there were mono-mers, e.g. AAAA -> AAAAA, it can mess up over the number of bases in a mono-mer.
  • PCR amplification can also cause slippage against monomers, same thing as the above point - but it's real.
  • 1 vs 0 ... this is because in a Pythonic list or any C/Perl-based array numbering starts at 0 (not 1) - hence @terdon's comments.

If upstream mono-mers are present that might be the source of the confusion here and given the amount of sequence data thats very likely.

What I'd do:

  • Just Blast the new sequence and see what emerges: if it is genuinely novel - they'll be zero 100% identity hits.
  • VCFs ... I think this is VarSome's stuff (forget).

If there are Blast hits it will identify whether there is any upstream lab artefacts.

Conclusion If this was any other situation in human genetics, or even mammalian genetics I'd be excited. However, in this case - I think its a routine result. Firstly, the chance of finding a new "driver" mutation are slim. Secondly, I've looked at this chromosome in a cancer patient (only once) and it was nuts (and thats an under statement) - I've a fairly good memory for numerical patterns, and I'd never seen a pathogenic species behaving like that. I think this insert is just a consequence of the entire mutational chaos that are cancer cells.

Just to make it clear: I don't do cancer but if you want an authoritative assessment and think you've something major I'd contact (or your PI) the Broad Institute via Geraldine A. Van der Auwera who also has an account on this site via @Geraldine_VdAuwera. Sanger sequencing is not what they do, but they'd know at the drop of a hat what the significance and if they don't respond - thats your answer ;-). Alternatively, VarSome would know, albeit they're commercial enterprise offering freeware.

  • $\begingroup$ @terdon I have the reads generated by sanger. If there is a tool I can use to get the insertion coordinates to confirm? The sanger generated results are doubtful because according to them the protein terminates due to insertion between a codon that would indicate severe condion which is not suggested in the patient's history. $\endgroup$
    – user16732
    Jan 6, 2023 at 13:41
  • $\begingroup$ The insertion is GAA so is it possible that the sequence shifted the results one base ahead? I have a picture of graph if you want to have alook $\endgroup$
    – user16732
    Jan 6, 2023 at 13:44
  • 1
    $\begingroup$ One thing you can try, @sanaamir, (and, again, I am pushing my own company's product here; I'm sorry, but those are the tools I am most familiar with) simply pasting the read into VarSome. That will align it against hg19 or hg38 (your choice), and identify any variants found in that read. That will help you see the full notation of the variant. $\endgroup$
    – terdon
    Jan 6, 2023 at 14:07
  • $\begingroup$ Thank you I will do that. But do i put the files from sanger or just fasta format of nucleotides with the insertion? I did have a look at the website today. @terdon $\endgroup$
    – user16732
    Jan 6, 2023 at 14:10
  • $\begingroup$ Please come into the chat room linked to the question (chat.stackexchange.com/rooms/141809/…) so we don't spam M_. $\endgroup$
    – terdon
    Jan 6, 2023 at 14:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.