A Scientific Quest for Community — Matchmaking and Networking in the Realm of Life Sciences Innovation
I am the co-founder of scifind.net, the Life Sciences collaboration network. I am a scientist with a background in molecular biology and genetics. I used to work primarily in genomics, with a focus on mtDNA and immunorepertoire sequencing. I went through the second cohort of IndieBio with the startup Girihlet Inc as their first employee. Through my experience, I saw lots of startups emerge from the ground up and witnessed their evolution at both IndieBio and later, the Oakland Genomics Center.
Now, science is tough–real tough. The sheer complexity and granularity that can be available in any particular subject is mind-blowing. Take Genomics for example–or well, let’s even go into sequencing. Then, deeper, into say RNA sequencing. From that it deviates more; is it microRNA sequencing, smRNA, circular, TCR?–the list goes on and on. Even within those subsets, there are more complexities, and we haven’t even gotten to the analytical portion.
There are library preparations, bioinformatics pipelines, all the different kinds of machines and protocols and reagents that can be used. Just that little overview should induce a bit of anxiety. Now, it’s not meant to be scary, only a glimpse into an ever-growing universe. Compound all of that by the thousands now, and you have an idea of the research industry in Life Sciences.
The way in which connections are made in Life Sciences are traditional (to say the least) compared to other industries. There is a high focus, both temporally and budgetarily on attending conferences. This is where scientists disseminate information, meet with one another, discuss research that they have not published yet, and much more. Although labs collaborate with one another, the work of a scientist is all-consuming, narrowly focused, and can be typically insulated within their own cluster. Unlike other industries where social media digital networking reigns supreme; life sciences has a focus on word-of-mouth, trust networks, meritocracy, credibility. Though these methods are powerful and indicative, they are also sluggish. To have word-of-mouth, predicates that you have the network to begin with, enabling a bit of a chicken or the egg scenario. To make it work, someone–or something–has to play matchmaker.
As a young scientist, going to conferences felt like a relic maintained by the old guard. The work is definitely interesting, but the execution of it all is stuffy at best. The interactivity and joy can feel virtually non-existent especially for an upstart. Communicating in a room full of strangers without any idea as to who they are ahead of time is peculiar, who exactly are you supposed to go up to and talk to? There is an almost driving insistence that all the big deals and collaborations are predicated upon relationships–that these identities and relationships can only be built with methods opposite to the internet. There is certainly a benefit to both, but for digitally integrated individuals, there is a comfort and expediency in online interaction–a medium that is lacking in this space. Take Instagram’s direct messages, for example, although it can be inundated with inquiries, many people swear by the collaborations and connections they establish through it.
With the advent of COVID-19, all of these typical communication channels have been disrupted. Without conferences, networking has come to a grinding halt. If you didn’t have existing network channels, they will be even harder to develop without the proper digital infrastructure to support new pathways of communication.
To top it all off, knowing if someone is even relevant to speak to is not that obvious. In order to profile scientists correctly, one would have to read many scientific papers to gain an understanding of their work (to determine pertinence prior to outreach). After that, there is an entire cold outreach process one can undertake, a process that most people do not like indulging themselves in.
The strangest thing about the scientific world is the way in which its members communicate and establish their self-organized networks (we’ll get to the weirdness of that). Current methods of networking are decentralized, and lack in overall unity. It could be a forum about specific topics, a Reddit group maybe, but there is little cohesion in the existing systems. What are the other options for startup owners and academics?
Despite the fact that Life Sciences’ is amazingly innovative in optimizing its manufacturing and research (as evidenced by the swiftness in developing a COVID vaccine)–its self-organizing framework around communication makes it hard to pull in information and discussions from multiple sources.
I’m a millennial, a user of social media, and I’ve been an avid gamer all my life (If I told you how many hours I’ve sunken over my lifetime into World of Warcraft you would be appalled). As an individual who has been digitally plugged in since I was old enough to speak, it’s bizarre that a lot of this network missing.
For a field in which discussion is integral, communication is decentralized in the Life Sciences. Anytime questions arose, it was almost as if you had to newly synthesize a community in some aspect. Practical questions could be asked in Research Gate or Quora, or outreach to lab mates and colleagues for advice or recommendations on a matter. There still was something missing there. Determining pertinence was still difficult! Why couldn’t people be clustered by their skillsets, their abilities? Like equipment in an MMORPG, there is a granularity to the stats afforded to an avatar by it that is akin to the honed technical skillsets of scientists. In the online scientific community, the sharing of information can seem transactional or cold. What if there was a way to band them together, to align scientists like the researchers of old?
I believe that AI is the answer to a lot of these problems.
We can circle back to the original problem of trying to prospect scientists. To do so, you have to read all their work and generate these profiles. The work is usually significantly contextual and understanding it (and assigning relevance) requires a certain level of scientific literacy.
Now, AI has been implemented in a lot of ways in science. From assisting in imaging diagnostics with identifying tumors, to drug development in which it helps predict what a potential blockbuster compound could be before sinking hundreds of millions of dollars into development. Additionally, with torrential quantities of digital health information such as wearables (FitBit or Apple Watch), medical records, sequencing data, and many more sources–utilizing AI to parse through it all is of utmost importance in making data-driven conclusions.
The space in which AI hasn’t been harnessed to its full potential in Life Sciences? Networking and communication.
The thing about science, compared to other industries, is that the industry pushes towards publications. This means that any scientist (or at least, any scientist worth their salt) has a vested interest in publishing and making it known what they are working on (to an extent). This means that for every individual, there are vast amounts of information that pertain to the work they do that is available for everyone to see. The dilemma? It all has to be read.
By utilizing AI, one could read papers and classify scientists. The use in that? Well, everything from building a network, finding expert consultants, recruiting, collaboration can be streamlined with such tooling. Instead of having to dig through the millions of papers and information floating around on the web, AI can consolidate it into palatable formats.
One of the fascinating features of doing something like that is that science becomes kind of like a dating application. It’s best to give a more concrete example of how it could work.
“With such huge levels of specificity in their fields, knowing whether someone is worth talking to can be tough–knowing involves stumbling upon them.”
You’re a young and aspiring microbiologist studying tuberculosis. In your area of research, you spend time culturing Mycobacterium tuberculosis (the infectious agent behind Tuberculosis) and perhaps administering different compounds to it in vitro to see what eliminates it. Elsewhere, a bioinformatician is studying expression data from tuberculosis experiments, poring through analytics. In another part of the world, a researcher is infecting mice with M. tuberculosis and testing novel compounds or seeing expression data in mouse lung tissue afterward.
What if they all sat in a room together and discussed it? What would that take in the current world? Well, there are a few avenues. By pure happenstance, they may find themselves at the same conference on TB, but perhaps the bioinformatician only goes to bioinformatics conferences, the microbiologist to microbiology ones? When and where do they meet? Maybe a friend introduces them but all of this hinges on the fact that they even know what to ask each other. A microbiologist, being so embedded in his field, might not know the pertinent way of discussing this with a bioinformatician. With such huge levels of specificity in their fields, knowing whether someone is worth talking to can be tough–knowing involves stumbling upon them.
Yet another factor is chance. Why leave it to chance? If AI can read things at an accelerated rate, it can provide people–not unlike a dating app–that you should network with and break the ice. In a way, this transcends current barriers that revolve around word-of-mouth, expensive conferences, and knowing what and who to ask.
Where is the intersection of science and socialization? I really started to delve into this question when I met my co-founder Stefani Robnett. She had been working at Facebook and naturally looked at the world from a social network-oriented lens. She loves political theory, interconnectivity, all topics which at the time were foreign to me as a traditionalist in the ways of science. We came together to work on developing a solution for communication in science and we’ve been on that path for the past year with our company, Sci Find.
I like to think of myself as an extroverted person but when I put that scientist mask on, I suddenly become a total weirdo. I’m not the only scientist with a penchant for insulation, obsessiveness over their research, curiosity, learning. We’re all a bit hooked on knowledge. There’s a certain indulgent exclusivity when you’ve been working on an experiment, you make a breakthrough, and for that brief moment in time, you’re the only one who knows this truth in the world. Those traits are double-edged swords but they don’t stop us from yearning for community, collaboration, conversation. The path of devotion can be a lonely one.
LinkedIn was genius in creating their professional platform. It was the first digital approach to creating a professional social network. Industry professionals swear by it but it offers less utility for the emerging scientist. As an amazing horizontal platform, it struggles in certain verticals; Life Sciences is one of them. People might be using it to share content, and perhaps even opinions but I don’t see it as a tool for solving scientific research problems. LinkedIn is bereft of a particular contextuality that must permeate scientific endeavors. It’s a one-size-fits-all for a highly complex and niche industry.
The easiest way to explain the problem is with a simple exercise.
If you want to try it out, type “tumor microenvironment” into Google. You’ll see tons of academic articles and information which is great, but it’s not pointing you towards people and companies.
If you try the same thing on LinkedIn you’ll get a list of people who are currently doing some work in the field. The exposition they give can range from minimal to substantial and it’s a mixed bag–it’s contingent upon the user disclosing the information.
If you try CrunchBase you’ll get a list of companies that explicitly say they work on the subject. These are all avenues that solve different problems, but none of them are speaking with each other and are prone to error. CrunchBase is limited in its quantity, Google is scholarly and LinkedIn is inundated and loosely categorized.
Companies do not verbalize all the things they do–you have to look at the people who work there to know. People, on the other hand, have tons of expertise, so without a look at the company, you don’t know what they are currently working on.
Take TB microbiologist from before as an example. They’ve got lots of technical expertise working with mycobacteria (and bacteria as a whole) so chances are that’s what they will be working on regardless of which company they are in. However, to know the topicality of it, you would have to know the company’s inner workings. If they are working on infectious diseases but TB is not in their portfolio, then it’s probably something else. Ultimately. there is an intersectionality between these entities that is paramount in science.
Social media platforms have become eerily great at predicting my personal tastes for consumer products. At this stage, they have lots of data on me, so they know I love sparkling tea drinks (among other things). However, on LinkedIn, despite having mass information on my connections, my work experience, publications, and more, still fails to realize that as an NGS expert there is little need for antibody sequencing in a genomics lab. This means I have little trust in its ability to refer me to things. The ability to match is integral in expediting innovation.
Broad search engines, like Google, are experiencing a decline in usage. As Vincent Granville says in his article that diversification of search engines into more niche and targeted approaches is an enormous opportunity. Think about some of your own consumer decisions. When you decide to look for houses, you go to Zillow; for household items–Amazon; CarGuru for Cars; Airbnb for rentals. You don’t use Google for any of those things. The same can be said for scientific exploration. The rise of vertical search engines as a competitor is essential in optimizing business processes. Other industries can get away with broad net approaches, but science itself is built upon immense granularity.
OkCupid was a pioneer in dating by creating a highly data-driven model of matching people (on even the most esoteric things). Now there comes a time where the same power can be harnessed for innovation. The new scientist bridges these powerful social channels, they communicate in ways natural to the zeitgeist. The conference will always be there, the lab will always be there–but now the conference is digitalized, the lab is the cloud, the world is your network.
Emerging scientists and innovators are ready for a new way to do business. They are ready to be recognized, to unify, and to enjoy the benefits of a global unified network. They are ready to push their innovations. Like most of the greatest advancements in the world, they almost exclusively come from the bottom up. Revolutionary scientific innovations start on humble desks, and in a world that tends increasingly towards industry, it’s important to not lose sight of where it all begins. There cannot be true scientific progress without adhering to that grassroots idea–that glimmer of truth, that scientific spirit.
