Transcript

JMP for data management in pharma

ANNE MILLEY: You co-authored a paper, Building Process Understanding for Vaccine Manufacturing Using Data Mining. Could you share some of the challenges and things that you explored in the writing of that paper?

JULIA O'NEILL: Yes, that paper actually describes the problem I've been mentioning from about 10 years ago. I'd say the biggest challenge happened before I took leadership of the project, and that was collecting all the data. And it's a very common problem in pharmaceuticals, even today, even with brand new companies, that all the key information is still on paper. My view of that is that that's actually probably the right thing to do for at least the first year or so until a process is really well-established and settled, and then lock it in. That's just my opinion.

But anyway, it was about 1,500 hours of time to gather the batch records and enter the key information into a database. It was a huge effort. Once we had the data electronic, JMP was incredibly helpful in all the tools for data management and organization. And I find myself, quite often, just begging people to just give me the Excel sheets they have, because I know that in JMP, within a maximum of an hour or two, I can have the data ready for analysis, all cleaned up, transposed, subsets, split, whatever. I love character formulas. I have all kinds of tricks I use. I often will have people that spend three days, a week, their weekends, trying to fix all those little things in Excel.

I know I could have done it, and I push and push. When I'm with people, I'll show them JMP and give them a flavor for, oh, you know, if you—

ANNE MILLEY: You'd save so much time.

JULIA O'NEILL: Yeah, and so gradually over time, a lot of the scientists are really loving JMP. It's catching on for sure.

ANNE MILLEY: We continue to try to remove those impediments, because you want to get to the fun part, where you get the insight.

JULIA O'NEILL: Exactly.

ANNE MILLEY: And you have to be familiar with the data to know what shape it needs to be in, but anything we can do to make that quicker and less painful.

JULIA O'NEILL: Yeah.

ANNE MILLEY: Good. Well, thank you for advocating.

JULIA O'NEILL: Oh, sure. I don't know what I would do or how I would do my job without JMP, and a large part of that is actually because of the data management aspects.