How to identify chromosomal translocation or deletions?

[spng]

I have a mouse cell line, which could have large-fragment deletion or translocation. I thought that the most affordable, efficient way to identify the translocation/deletion loci could be exome sequencing; at least I can identify which genes are lost in the mouse genome. I am not an expert on genetics or sequencing, so any comments or thoughts would be appreciated. Also, I welcome sequencing vendors to provide quotes about the possible project. Thanks.

[Brian Bushnell]

If a large-scale event is heterozygous, I don't think exon-capture would be particularly useful; it does not give useful information about copy count because of the bias in bait efficiency. You MAY get some reads spanning a deletion, but then again, you may not; depends on whether the boundaries fall within the baited region. Also, some allegedly targeted exons will just not show up even though they are present.

The cheapest way to look for large-scale CNVs is with an array chip. It won't give exact boundaries, but with a 3Gbp genome, a 1 million SNP chip should give a resolution down to around 3000bp. Also, this won't tell you anything about a copy-neutral translocation.

Whole-genome sequencing is probably the most reliable method for finding novel junctions from translocations and deletions, though even then, discerning true translocations in large, repetitive genomes requires a substantial amount of skilled manual labor.

[Joann]

First of all, how stable is the karyotype?

[spng]

Brian, thanks for your helpful suggestions. For my project I wish to get a clue about the genes responsible for the phenotype and then I will work on the gene candidates using wet lab techniques to rescue or recapitulate the phenotype. So I may not be so interested in the boundaries. I was also thinking about using RNA Seq, but that is not a direct and reliable approach to my aim, although I may get some clues about the downstream signaling. So I agree that your suggestion of using array chips is a better idea. And, I guess that array can work with heterozygous events (please correct me if not).
To Joann: about the karyotype, I have to say I did not look into that; so it is good to know about this.

[Joann]

Right, at some point I am hoping NGS will provide a simple assay mechanism to monitor these things that happen under tissue culture conditions like changes in chromosome number, which is a pretty macro event as well as other even harder to visually observe changes in chromosome structure, both qualitative and quantitative. In order to better validate results obtained with tissue culture models as well as to better understand cell division dynamics under cultured/immortalized conditions.

[Brian Bushnell]

And, I guess that array can work with heterozygous events (please correct me if not).

Arrays data I've worked with was able to indicate copy-count, since it's basically an analog measure of signal strength. The biggest disadvantage of arrays is that they are expensive and time-consuming to initially design and create, but for mouse they should already exist.