«PrevChikungunya, a Neglected Tropical Disease, has Re-emerged as a Global Health Risk NextOutsourcing your patent translations – the intelligent choice»

Donald Johnston

Download the Full Interview

As William T. Loging sees it, bioinformatics – the harnessing of algorithms, large databases and other powerful computerized tools in the study of biological processes – offers benefits beyond drug discovery (or, as Loging prefers to phrase it, drug “engineering”). In this interview, conducted by Donald Johnston, Senior Director, Current Awareness at Clarivate Analytics, Loging describes how computational tools can help a pharma company focus its efforts, allocate resources and find the best way forward.

Loging is the editor of Bioinformatics and Computational Biology in Drug Discovery and Development (Cambridge University Press, 2016). Trained as a molecular biologist, and a former research scientist at both Pfizer and Boehringer Ingelheim, Loging is currently associate professor of genetics and genomic sciences and head of production bioinformatics at the Icahn School of Medicine at Mount Sinai in New York.

The Q&A below is excerpted from a longer interview

Q: You’ve suggested that the term “drug discovery” is a misnomer. Explain what you mean by that.

Loging: Back when Alexander Fleming “discovered” penicillin, that was pure discovery. He wasn’t necessarily looking for anything other than why bacteria weren’t growing in a dish. But he based his finding on an observation.

Today, we are not discovering drugs – it’s more of an engineering process. That is, we’re taking the learnings from the discovery of penicillin and we’re going, “Okay, let’s see how we might alleviate the symptoms of this disease. This protein is what’s causing the problem.” If you could somehow modulate that activity then you could “create” a drug. So it’s not necessarily discovery, because the small-molecule panels you screen already exist. The focus started out primarily on small molecules, but now also includes bio-therapeutics – and it’s the same approach with antibodies.

After coming up with an entity that will modulate that protein’s activity, some people may say, we’ve “discovered” it. I personally say we really “engineered” it, because we went into this space knowing what needed to be created. It’s very much like engineering.

Savvy business people sat down and said, “Wow, scientists can create drugs. And, 1) those drugs can help people (which is very important); and, 2) we can make money on the approach. Why don’t we engineer serendipity to move this process forward?” So if scientists take an idea to address the cause of a disease, which is very often a protein, and work to modulate its activity, then we might be able to alleviate the symptoms of that disease. Everybody talks about identifying a target. That process varies wildly between different companies. Sometimes the idea comes from reading a scientific paper [or observing patients]. For example, the drugs that target proteins such as CCR5 [C-C chemokine receptor type 5] or PCSK9 [proprotein convertase subtilisin/kexin type 9] started in very much the same way – with observations of patients.

Q: Money is a constant research challenge. Companies have to make R&D budget choices. How can these computational tools help identify which asset is worth driving first?

Loging: Where we’ve shown success from a computational biology standpoint is being able to address just those answers. Historically, the tools have been used heavily to identify new targets. But, you know, 98% of all new target approaches fail. So people are leery about making big investments in that space. They are more willing to use these approaches toward the end of the pipeline. So, for example, if I said to you, “If you invest a million dollars in this process, I’d be able to open up other indication areas that could potentially lead to about a $50 million increase in revenue year over year at the very minimum for this particular indication.” That’s huge! Who wouldn’t want to do that? So from our standpoint, we really like the idea of doing “indications discovery.”

I remember very clearly in 2008 while speaking with several colleagues that I wanted to jump-start existing drug programs into other disease fields. And a lot of the response I got was, “I know what indications my drug candidate is for. I know which ones are working.” Those responses are very short-sighted. Today, computational biology can provide a lot of value in being able to identify the indications for a specific drug candidate or its target.

Also, if you sat down with people from the marketing and business area, you’d explore questions like, “Of those indications, which ones might we be able to move into? What’s the competitor space look like? How much are the competitors working on this?” Those are all the areas where computational biology can provide a very high return on investment.

Q: Are you saying that bioinformatics and computational biology can help set priorities, helping management decide the next step?

Loging: Very much so. If a company plans to start a clinical trial, it’s very important to understand how the drug candidate compares to any standards of care or differentiates from competitors. And it’s also being able to communicate with them effectively and saying, “Here’s the next experiment that you can do.” That opens up people’s eyes. It has to do with the fact that I have worked at the bench for several years. I am seeing it with both viewpoints. Many times, in silico scientists have taken data and presented it to their management – or to other people – and they’ve said “Here you go,” and the management response is, “I don’t understand this. This information is not useful.” I find that it’s not enough to just do the analysis, but you also have to message it correctly. Very often, computationalists have to go that extra step. They’ve got to also be able to say, “Oh, here is what you should do next,” being able to connect the dots for people. It’s almost like an art form because you’re saying, “Here is the data and here is what you should do with that information,” but you don’t often see that type of approach or skill.

I’ve purposely sought out people who aren’t in our field, in order to speak with someone with an outside perspective and understand how they sell their idea. I can’t overstate how important this communication and the “relaying of ideas” process is for the pharmaceutical industry. It’s by no means dumbing things down – it’s just translating information so people can better understand the findings.

Q: Investors, partners, potential acquirers, they are all looking for “differentiation.” How can bioinformatics help you get there?

Loging: This is the exciting part. The current technology is allowing us to really understand how drugs affect individuals at the molecular level. Even drugs from the same class will have very different effects within a human cell or within an individual. Bioinformatics allows us to take that entity or that drug candidate and say, “Okay, I am going to compare every change that happens from a functional standpoint.” And then you start looking through that data.

It’s affecting processes like cholesterol metabolism. It’s affecting cell mobility. It’s affecting a wide array of cellular functions that may or may not be related to your disease. By applying statistics you can rank-order those. You can ask, “What is the most significant function that is changing?” We’ve done that, and we’ve published some of these things in the past. We’ve been able to show that you can compare to the current market standard of care, taking all those functions that change, and then aligning those – it’s almost classical comparative biology – aligning that to your candidate.

For the complete interview with William T. Loging, click here.