Smart Tables: Analyzing Oncology Trial Design and Endpoints with Maven Bio 🔬

My name is Arjun.

I'm the co founder of Maven Bio.

Today I'm going to walk you through a quick example of how to use our Smart Tables feature to analyze a set of oncology trials, looking at their patient population and their endpoints.

So let's get started.

First, we'll navigate to the Smart Tables component of the Maven platform.

You have the ability to import a table from Excel that has a list of drugs or trials that are of interest and then you can use Maven's AI on top.

Or you can create your own search within Maven, which is what we'll do today because I'm going to be looking at trials specifically.

I'll update the search type to that and then I'll type in my request to the filter suggestion box.

So I'm going to say phase three trials for prostate cancer completed in the last five years.

You also have the ability on the left hand side to directly select the filters.

But I often find that the filter suggested box works pretty well in most cases.

So we take a look at what it's suggested that matches my intent.

I'll apply those filters.

Now, once we get this list to run our AI columns and do further enrichment, we need to first create a Smart table on the right hand side.

What that allows us to do is save those results.

I'm going to say prostate cancer phase three test.

Now, all the standard columns are already available and you can add more of them by navigating to the column window here.

For example, if I wanted to know the enrollment, click the button and it's added.

But let's say we wanted to understand something that wasn't in those standard columns, maybe around patient population and trial design and most importantly to get some endpoints that that's where the AI columns can help us do that work.

So what we'll do is use natural language to write the additional questions we want to add to that table.

So let's take a look.

First, determine if it is a basket trial basket, defined as studying more than one type of cancer.

Second, determine if the trial is focused on metastatic prostate cancer specifically.

Third, find orr, and PFS for prostate cancer.

Focus on primary endpoint timeline.

So I've had a couple distinct requests there.

It'll ask me some clarification questions and then it'll convert the prompt into a preview table so that we have a sense of what it's going to add before it starts running.

That gives us the opportunity to make any changes or edits needed before it gets going so let's take a look.

This matches what I was looking to do and if needed I can delete or edit columns.

But I like that it added the context and the timing for us here because that was something I'd probably want to know anyways, but forgot to write it in.

And one thing you can always remember is you can save these as a blueprint which will then be available in the blueprint library on the left hand side so that when you come back to it you can run that same set again.

Once we apply them, it'll start to add those columns to the table and create those AI columns and use dozens of AI agents at once to enrich this list and get the information we need in a structured way.

This can be very accelerating and has a wide range of use cases.

So here we've done some around the trial design, some around the patient population, some endpoints, but you can also use it for a wider variety of things on Drugs, on companies, etc.

Very powerful tool.

So hopefully this was helpful for you today.

Thanks for watching and take care.