Learn from The University of Iowa

The University of Iowa Bioinformatics Division was created about 10 years ago to bring together under one organizational umbrella expertise in human genetics, clinical genetics and sequencing, and educational outreach. The goal was to enhance the ability of the hospital to do genetic outreach to the people of the state of Iowa, to improve the ability of the hospital to provide genetic counseling services, and also develop clinical genetic tests.

The Bioinformatics Division provides support for analysis of sequencing data, clinical genetic data. It collaborates with laboratories all over campus, but especially in the Carver College of Medicine to get data analyzed and help researchers make sense of their sequencing data, and publish papers. The core facility performs hundreds of pathway analysis runs per year.

Michael Chimenti, the head of the division, did his PhD at Johns Hopkins in biophysics studying protein structure and function. He then went to UCSF, the University of California San Francisco, to the pharmaceutical chemistry department, and did a postdoc in pharmaceutical drug design and computational chemistry, worked on small molecule drug discovery. And then, from there, leveraged his computational expertise to come back to, where he got my bachelor’s degree here, and began performing bioinformatics analysis. He started at the associate level, and then in 2019 was promoted to director of the group.

Advaita was recommended to Michael by the genomic sequencing director. He had had some contact with Advaita and he told Michael that, “You might want to take a look at this. It could be interesting.” Michael says: ”Advaita wasn’t on my radar prior to that, so it was his recommendation to look into Advaita more deeply and schedule a meeting. I had already had some growing dissatisfaction with IPA and the way IPA worked. And so, I was primed to look at alternatives already. And then, when director suggested that we check it out and talk to someone, I was open to that. That meeting with the Advaita rep went and I was immediately impressed by what I saw.”

Michael states that his dissatisfaction with IPA was along several lines. He felt that the pricing was exorbitant. He says:” From my perspective as a client, I saw the institute spending $50,000 for two seats for a single year. And, it seemed excessively expensive for a piece of software. And, helping people do their analyses and delivering results was such a chore, because it flooded the investigator with so many results pages, and so many diagrams, and figures, and things to decipher that it was making more work for me. It wasn’t solving problems; it was creating problems. And a lot of it was not really interpretable. I mean, it was just results lists after results lists. I just haven’t found it very interesting. I just found it unpleasant to interact with, I would say. It just was too much going on, too busy, you couldn’t get to what mattered fast enough. And, as a core director, it created more problems for me than it solved.”

Michael says that the switch to Advaita was an easy decision and an easy conversion. He believes that because of the cost of IPA and the ongoing lists of tables in IPA that he has turned his core from a cost center that was losing money into a profit center that provides significantly better value to researchers. The researchers at Iowa had also tried DAVID. Michael states that: “There are cases where I have clients tell me, “Wow, we struggled with this gene list, putting it into DAVID, putting it into free online tools, gene ontology, explorers, and things, and never really got anything. And then, you did this iPathwayGuide™ analysis for us. And wow, the top hit was exactly the pathway we’re interested in. We figured it was important, but here it is showing up in this report.” I’ve heard that many times from people.”

Michael adds: “iPathWayGuide is doing pathway analysis with a high degree of scientific rigor. So, nothing’s being massaged, or P-hacking, or manipulation, it’s just, because it’s a better algorithm, it gets you to the answer faster. And, that’s perfect.”

Michael says that the license with Advaita was structured put him in charge as a core director, rather than disempowering him. He adds: “So, with IPA, the institute was buying the seats and that expense was showing up as a cost in my budget. So, it was driving my division’s budget deep into the red. And then, we were giving away access to it for free to the university. And so, it was a huge cost center that made my budget look terrible. But I didn’t know who was using it. I would just get questions from people like, “Hey, I need help with this IPA analysis.” Or, “Can you make sense of this graph or this table?” Or, people send me gene lists. And then, they would say, “Can you do this analysis?” And so, I’d have to wait for a seat to come available. And, it was just a headache and just not helpful.”

And with iPathwayGuide™, the way that the license is set up, the fact that I and two of my staff can actually generate the reports, turns it from a liability to an asset. So now, it’s a selling point. It’s, “Work with us. We have access to the best pathway software. We will make this for you. And, you don’t have to wait for a seat. We’ll just make it right now. And then, you can look at it anytime you want. And you can share it to whomever you want. And they can look at it anytime they want. They don’t have to wait for a seat either.” I mean, it’s brilliant from that perspective. And, like I said, it gives us an asset, it gives us a feature that we can sell.

Michael believes that from his perspective as a core director, that is just another value add that he can bring now. The license cost is reasonable, he can fund it through charging for reports at a reasonable fee that people will willingly pay. He believes that based on our volume of reports, this covers the license cost. It pays for itself, it’s a value add and people want to work with him because of that. And so, that’s why he likes it.

He adds: “When I show them and I teach them how to use the network analysis module and create their own custom networks from their results, that’s an eye-opening moment for a lot of people, where they say, “Wow. This is the insight that I needed. This is the figure that I’ve been wanting to generate where I take my gene list or my particular pathway of interest and generate this network that can be rearranged, it can be pulled, and stretched.” That is pretty powerful. And then, you can add chemicals that interact with it. You can add upstream regulators, all these things that you can very quickly get to a plot that allows you to see your results in a network form that is insightful and meaningful, and not just a huge pathway diagram that just overwhelms people.”

He goes on to say: “Because I’ve seen a lot of papers with IPA results with gigantic pathway diagrams with a few things lit up here and there. And, what does that mean? I mean, it’s impossible to interpret, but with the network analysis in iPathwayGuide™, you can start with just your collection of genes that are actually measured, that are differentially expressed in the network, build a network around them by going one hop out, one jump out to known associates, so to speak, known molecules that interact. And then, you can have different topologies, and end up with a figure that tells a story that you want that makes sense.”

Michael also likes the shareability of results in iPathwayGuide™. He says: “When I send reports to a someone, they’ll share it within the lab to other postdocs or to the PI. The network plot results in figures and papers, and is quite beneficial from that perspective, because instead of just having a table that lists pathways that were changed, having this overly complex generic pathway plot with a few things highlighted, this is actually a custom network that describes the biology that’s relevant to that study. And that makes a huge difference. So that’s another feature that I really like about it. I also like the ability to download a PDF that summarizes all of the results from the analysis, including the methods and all of that in one large document. I like that the help is built into the interface, so you can see the videos that describe how to use each section. And, that you can export figures from it pretty easily.”

Michael adds that he gets many positive comments from his users. He says that many users tell him that the iPathwayGuide™ reports get them to relevant results quickly, and help them make sense of gene lists and differentially expressed gene lists when other tools did not. And the proof is in the pudding, in terms of a lot of papers that they published with iPathwayGuide™ results in the fact that papers that are almost completely based on iPathwayGuide™ reports that all the figures are from iPathwayGuide™ reports. He adds: “That’s the ultimate proof of value added to the analysis, to the experiment, is that, you’re publishing papers with exactly those results.”

When asked what he would say as a reference Michael says: “That iPathwayGuide™ is a cutting-edge tool for doing pathway analysis that very quickly gets you to relevant answers, and accelerates research, and publication. It is easy to interact with the reports. The reports are in a convenient online format and easy to share. And, it puts researchers and subject matter experts in the driver’s seat of interpreting their own data and takes that burden off of the bioinformatician. Yes, of course I would certainly recommend it from a cost, technological superiority and usability perspective. It’s an ideal solution for a core like ours.”