DNA sequencing from cord blood stem cells

Question

[dodausp]

I believe this is more of a broad question; and I am resorting to here because I found very limited (and not specific) information on the topic. Briefly, I am planning on performing DNA sequencing in mesenchymal stem cells (MSC) extracted from cord blood, with the following in mind:

1. Are those cells a good source (in terms of quantity and quality) for germline SNP's investigation?

2. If so, do the discovered variants truly reflect the presence of a germline variant?

What is known about mesenchymal stem cells is that they can be quite heterogeneous and harvest site-dependent. Specifically, for cord blood-derived MSC the heterogeneous composition was shown to undergo dynamic clonal selection in culture, with an initial massive reduction in diversity and a selection of single clones over time, starting in the early passages (Selich et al., 2016).

So, the most concerning questions I have for now are:

1. Is it a reasonable approach to use these MSC's for germline variant investigation? (considering that (a) the MSC's have been expanded for 1-2 passages prior to sequencing; and (b) it is a panel sequencing rather than WES or WGS)?**

2. What would you opt as your analysis pipeline, GATK best practices?

3. What else should I be considering in this experimental setting?

Any help is much appreciated. And of course, article references are also very welcomed. Maybe this is a topic with a lot of discussion within the research field, and I might be missing it. I don't know.

Thanks a lot!

Answer 1

[mark.ziemann]

1. Is it a reasonable approach to use these MSC's for germline variant investigation? (considering that (a) the MSC's have been expanded for 1-2 passages prior to sequencing; and (b) it is a panel sequencing rather than WES or WGS)?**

Yes, basically any cell type will be fine for identifying germ line mutations. Even the cord blood mononuclear cells. What are you comparing to? Do you have parental samples too?

1. What would you opt as your analysis pipeline, GATK best practices?

Yes that seems standard for exome seq.

1. What else should I be considering in this experimental setting?

Hard to say, you haven't mentioned what the purpose of the experiment is and what comparisons you plan to make.

[dodausp]

Thanks a lot @mark.ziemann. And sorry for not delving much on the original questions; I was afraid to be very specific at first without getting into the core issue, hence not getting the interest for the discussion. So, thank you a lot for the feedback.

So, the overarching goal for this study is to address whether cord blood MSC's are good surrogates for hereditary or de novo germline conditions investigation (in the context of DNA sequencing), such as a handful of cancers or hereditary syndromes. It is, do those MSC's have the potential to be used for genetic counseling? Or even useful for a pre-screening method prior to a referral to a complete genetic counseling (i.e. involving parents and progeny)?

So, my points to your questions are:

Even the cord blood mononuclear cells. What are you comparing to? Do you have parental samples too?

Are the parental samples critical in this case? I'm asking this because my rationale is that by investigating the subject, both the hereditary germline variants (already present in the parents) and the novel germline variants (caused by germline cells mutation, recombination/rearrangements) would be picked up. Or is there a critical misconception I'm tripping on here?

Edit: also, I guess I misinterpreted the term "parental cell" here. Due to the context, I literally got as samples from the parents. Sorry about that. But if you meant as in "progenitor cell", wouldn't the mesenchymal stem cell itself suffice for that purpose?

Yes that seems standard for exome seq.

I am planning to use a gene panel for some hereditary conditions (i.e. ~300 targets) on this step. Here, thinking more about higher reading depth rather then genomic/exomic coverage. In that case, should starting off with the proprietary software analysis (for this study, IonTorrent) and finishing with a post-variant call "hard-filtering" pipeline (e.g. adapt the cut-offs set by the original analysis according to each run, and updating the ClinVar calls, and so forth)?

[mark.ziemann]

Parental samples are not critical, but can help you distinguish de novo mutations. It may also be important to identify whether the allele was inherited maternally or paternally for future management.