Gary Marcus

I've always found the majority of senior genomics researchers are in strong agreement (that it has all been a disappointment). I personally feel like we learned a lot about the problems with analysing massive datasets with tens of thousands of variables, and that is valuable.

The real question is do we double down with other omics (epigenomics, proteomics, microbiomics, etc.) or do we assume that all off this is nonsense, and if there were big effects (and targets) they would have been discovered by traditional science?

The title of the post should have been 'The HGP has not yielded any results in depression as yet'. Calling the entire project a 'dud' simply means the author has absolutely no idea how much ALL next-gen sequencing testing relies on the HGP today.

It's literally impossible to even know whether a patient has a pathogenic brca mutation or not without having the human reference genome. And then, of course there's the whole field of Onco targeted therapies. Very rare that I get worked up about random articles.

you may have a point but eg could you some of that with specific reference points (eg data on brca) rather than whole genome? how much (not a rhetorical question) does full genome buy you?

Not sure I get the question exactly, but here goes. Everytime you sequence the genome of a patient (for any reason, let's say ascertaining brca mutations) you need to compare what sequence you get against the human reference genome. This comparison gives you 'variants'.

These variants are then analysed to see whether they are disease causing (pathogenic) or not (benign). The entire genetic testing and targeted therapy industry relies on this.

So essentially, every human genome sequenced, for whatever purpose, anywhere on the planet, uses the reference human genome as a baseline.