All cells have the same genome and differ only by the expression pattern. Is it possible to determine the tissue origin of a cell based on its DNA sequence using short tandem repeat (STR) analysis or copy number variation (CNV) analysis?

  • 3
    $\begingroup$ You say yourself that all cells have the same sequence, so how could you tell them apart based solely on sequence? $\endgroup$ – canadianer Nov 16 '17 at 23:39
  • $\begingroup$ I don't know; that's why I am asking. Wikipedia page for STR analysis says that STR analysis can be used to track cancer cell progression, and I found a paper that claims CNV can occur in different tissues. I was wondering whether it would be possible to determine the origin of the cell using these or other similar molecular marker techniques. pnas.org/content/109/44/18018.full.pdf $\endgroup$ – John Smith Nov 16 '17 at 23:42
  • 1
    $\begingroup$ Let me rephrase my question as a statement: if all cells have the same sequence, you cannot differentiate them based on sequence. Cancer cells do not have the same sequence and thus could show STR expansion or CNV. I suppose you could also see mosaicism if such a mutation occurred early in development but I don't know the extent to which this occurs nor how strongly it would correlate with adult tissue types (other than in a broad classification such ecto-, meso-, endoderm). Too which paper are you referring? $\endgroup$ – canadianer Nov 17 '17 at 0:01
  • $\begingroup$ Oh I see the paper in your edited comment. Yes they seem to be referring to mosaicism and the variations they discovered appear to be between tissues that developed from the same progenitor. Interesting stuff. If you were trying to determine the tissue from which a cell originates, you may consider profiling gene expression, such as by RNA-Seq $\endgroup$ – canadianer Nov 17 '17 at 0:09
  • $\begingroup$ It's like looking at a binary code (101010101011110001110101) and asking "what would this do?" Actually binary is still far better than deciphering a genetic code... $\endgroup$ – bonCodigo Nov 17 '17 at 0:17

The answer is in the question but I suppose you wanted to ensure you were not missing some info.


The DNA sequence is the same in all cells of a multicellular organism. Only the expression pattern varies among tissues. It is therefore impossible to tell from the DNA sequence alone what tissue a cell come from.

It would be feasible by looking at the transcriptome or the proteome or eventually by looking at epigenetic modifications.

Potential exception

Of course, as you talk about CNV, if you consider that the number of chromosome is part of the info present in a DNA sequence, then haploid cells (spermatozoids and ovules) are an exception and you could tell them apart. Similarly anucleated cells would be an exception as they contain no nuclear DNA.

Further potential exceptions are explained by @tsttst in his comment.

  • 2
    $\begingroup$ This answer may be wrong as transcription can affect mutagenesis as shown recently by Sankar et al., Nature 2016, nature.com/articles/nature18316; while their great study was on bacteria, it could be thinkable that also in animals there is a coupling between transcription and mutagenesis (--> thus tissue-specific transcription profiles may lead to tissue-specific somatic mutation patterns --> which would allow identification of tissue by DNA); P.S.: A further exception would be en.wikipedia.org/wiki/V(D)J_recombination $\endgroup$ – tsttst Nov 17 '17 at 4:53

There was a paper in Science, August of this year (apologies to the author for not remembering their name). They worked up a protocol for single cell bisulfite sequencing and through the use of known marker genes managed to identify multiple neuronal sub-types based on non-CpG methylation across gene bodies.

Non-CpG methylation across gene bodies is seen often to be inversely correlated to gene expression, through this you can infer an expression profile of known marker genes giving you have a solid poke at a cell type with only a DNA sequence.

(Edit: Assuming this holds true in tissues outside of the cortex)


i would rather start with an assumption that no two cells in the human body have the same genome. Mutations occur frequently; most of them are repaired, but some remain unrepaired. Human body is a genetic mosaic.

Abyzov et al. (Genome Research, 2017):

We estimate that on average a fibroblast cell in children has 1035 mostly benign mosaic SNVs. (...) These findings reveal a large degree of somatic mosaicism in healthy human tissues, link de novo and cancer mutations to somatic mosaicism, and couple somatic mosaicism with cell proliferation.


  • $\begingroup$ Such mutations, being randomly driven, wouldn't help to identify which tissue the sample came from however. $\endgroup$ – Koen vd H Nov 17 '17 at 23:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.