[Locating, from scratch, the gene related to a disease is like] trying to find a burned-out light bulb in a house located somewhere between the East and West coasts without knowing the state, much less the town or street the house is on.

Are we using science in ways that it wasn't intended to, in which case we should be a little careful, or are we using faith in ways that faith wasn't really designed for? There are certain questions that are better answered by one approach than the other, and if you start mixing that up, then you end up in ... conflict.

To get our universe, with all of its potential for complexities or any kind of potential for any kind of life-form, everything has to be precisely defined on this knife edge of improbability. [Y]ou have to see the hands of a creator who set the parameters to be just so because the creator was interested in something a little more complicated than random particles.

Addressing the conclusions of The God Delusion point by point with the devastating insight of a molecular biologist turned theologian, Alister McGrath dismantles the argument that science should lead to atheism, and demonstrates instead that Dawkins has abandoned his much-cherished rationality to embrace an embittered manifesto of dogmatic atheist fundamentalism.

Research is so unpredictable. There are periods when nothing works and all your experiments are a disaster and all your hypotheses are wrong.

I had the analysis of a million or so SNPs [single nucleotide polymorphisms] just to see what was there. That's partly because I was writing a book about DNA and personalized medicine and I thought it would be a little bit disingenuous to talk about what could be done without actually having the experiment done on yourself.

When a drug comes out [that's broadly prescribed] there are going to start to be a lot of people on it [in a million person cohort] and you might get therefore an early signal of something unexpected that hadn't come through in the clinical trials. And I'm sure [drug companies] would love it if, in fact, FDA, recognizing that, would say, OK, maybe you don't have to do your trial with 30,000 people because we're going to find out shortly after registration because we'll have a lot of people taking the drug and we'll be able to see what happened using PMI.