It's all about making decisions
I’ve been writing about how to design a digital twin of a biotech lab, I’ve often referred to “the type of functionality that a biotech data team needs.” But I now want take a step back and understand what that functionality actually is, and why it’s necessary. So in this and the next few posts, I’m going to discuss the goals and challenges for different aspects and stages of a biotech research program, from the perspective of digital tools.
As we go along, I want to keep a focus on goals and outcomes, so I’m going to start at the end of the process and work backwards. And while there are different possibilities for what the end goal is, I’m going to take the example of preclinical drug development since that’s what I know. So I’ll assume the end goal is an IND (Investigational New Drug) filing - the report you file to the FDA to get approved to start clinical trials.
Now, the content of an IND is a complex subject that I am very much not an expert on. But for the sake of what I’m writing about, we should be able to boil it down to two things you need to have done to write the IND:
You need to have made a number of decisions about a compound/biologic/etc. to pursue, a disease indication and a number of other parameters related to the clinical trial. Since the next steps get very expensive, you need to be confident that these were the best possible decisions given the available data.
You need to present evidence that demonstrates how you came to these conclusions. Experts will have different ideas about exactly what this evidence needs to cover, but one way or another you’ll need trace the decision making process to show that it wasn’t arbitrary or biased.
It’s worth noting that the way you think about this will depend on how many INDs you expect to write. A drug program lead might expect to spend years on early discovery before they start to write a single IND. A data team supporting multiple drug programs may expect to be involved in multiple IND filings every year. I’m going to tend towards the latter, since we’re talking about data teams here, but each organization will have to make decisions based on their own expectations.
Either way, there are going to be three key themes as we work backwards from here: First you need to ensure the key decisions are made in a trustworthy and repeatable manner. Second, you need to make sure the best data is available to make those decisions. And third you need to document all of this in a way that you can find and explain when it’s time to write the IND.
In the next few posts, we’ll dig deeper into what this looks like throughout the process.