Published online 14 October 2010 | Nature | doi:10.1038/news.2010.540

News: Q&A

The renegade gene test

An open-source computer program flouts patents to test for cancer-causing gene mutations.

Steven SalzbergSteven Salzberg says patents on human genes don't make sense.

A freely available computer program can check any genome for 68 gene mutations that increase the risk of breast, ovarian and other cancers. Steven Salzberg and Mihaela Pertea of the Center for Bioinformatics and Computational Biology at the University of Maryland in College Park, who wrote the program, described it in a study published last week in Genome Biology1.

The mutations occur in tumour-suppressor genes BRCA1 and BRCA2, which have become notorious in the 15 years since they were discovered — not only because of the mutations, but also because of seven US patents on the genes that are held by Myriad Genetics, a molecular diagnostic company based in Salt Lake City, Utah.

A US court overturned the validity of those patents earlier this year — a decision that Myriad is now appealing. But regardless of the legal outcome, says Salzberg, times have changed, and such patents lock up information that should be accessible to anyone for free. His genome-analysis software is available with an open-source licence on the bioinformatics centre's website.

Salzberg told Nature about his plans to take the software project forward.

What was your motivation for developing the software?

It was a poke in the eye for the patent system. I haven't been particularly outspoken about it, but many scientists, including me, think that it doesn't make any sense to allow patents on human genes — they're not inventions. So the patent office had it wrong in the first place.

Human sequencing has become very cheap. I think that we will all have our genomes sequenced and on a flash drive in maybe 10 years. The point that we wanted to make was, if someone wants to take their own DNA sequence and check it to see whether they have a harmful mutation, why shouldn't they be allowed to do so? Patents get in the way. It is kind of a bizarre situation we're in — very expensive licenses are required to interrogate any mutation in BRCA1, and there are thousands of other genes that have intellectual property claims on them. So in a few years, when we all have our genomes available, you might have to pay thousands of different licence fees to look at your own genome! That is crazy!

The software includes just 68 of the more than 1,000 mutations in BRCA1 and BRCA2 that are known to have a role in increasing the risk of cancer. How hard would it be to expand your program?

It is grunt work. We have just submitted a proposal to expand the software to cover all cancer genes, and to go beyond that to all disease genes. We estimated that it would take 3 years, primarily for Pertea, with one student. It is not that big a problem. For a National Institutes of Health grant, the probable cost of about US$300,000 is not that much.

As we develop the software further, we're planning to add links that will point you to the paper on each mutation so that users can educate themselves. Now, most people won't understand the papers, but at least it is a start — you could take that to your doctor or somebody who better understands genetics, and you could say, OK, there is a paper about this mutation, do I have it and do I need to worry?

Do you have concerns about the legal repercussions of the software? How did that play into developing it?

We talked about that a lot. I'm not afraid of being sued — I think it is unlikely, because I don't think I'm a valuable target. I don't think that Myriad would gain anything by suing us. If they were to do it, it would probably be to make a statement.

Presumably you would have to stop your activities, at the very least.

What are we going to do? We've already released the software, and it is free. The great thing about open source is that once you release it and people start downloading it you can't take it back. I could take it off our website, but it is already out there.

How big can the impact of this software be, given that so few people have had their genomes sequenced?

Well, there are actually a couple of problems with it at the moment, and we acknowledge that. It only has 68 mutations, and Myriad's test covers many more. If you were worried about your BRCA1 and BRCA2 genes, you wouldn't use just our software.


But you're right — very few people have had their genomes sequenced anyway. However, we think it will take us a few years to expand the software to cover many more mutations and many more genes. By then, many more people are likely to have their genomes available. As we said in the paper in Genome Biology, at the moment the program is a proof of concept. The world isn't ready for it yet, the software is not ready for prime time, but we can see how to do it. We don't need any further breakthroughs to make the software cover all mutations. By the time it is fully ready for you to use and it is useful, then maybe it will cost less for you to sequence your own genome than it will to pay Myriad. 

  • References

    1. Salzberg, S. L. & Pertea, M. Genome Biol. 11, 404 (2010). | Article
Commenting is now closed.