Kanvas Bio Annual Science Day Recap


Kanvas Bio Annual Science Day Recap

by Kanvas Biosciences

Recap­ping Kan­vas Bio’s 2023 Sci­ence Day: A Q&A with our CTO and CSO 

Every Decem­ber, the Kan­vas Bio team hosts an in-depth review of our research and devel­op­ment accom­plish­ments from the past year. We also present our goals for the com­ing year. We call it Sci­ence Day, and it’s an impor­tant time for our team, part­ners and investors to col­lec­tive­ly reflect on Kan­vas Bio’s progress and future.

This year’s Sci­ence Day was filled with excit­ing updates that bring us clos­er to achiev­ing some of our biggest goals. Nine Kan­vas Bio team mem­bers pre­sent­ed, and we were espe­cial­ly proud to share the lat­est news on our lead drug can­di­date in our Immuno-oncol­o­gy Pro­gram, KAN-001. We spoke with two of the pre­sen­ters – Dr. Hao Shi, our Chief Tech­nol­o­gy Offi­cer, and Dr. Philip Burn­ham, our Chief Sci­ence Offi­cer – to get their per­spec­tive on the event, includ­ing what the major high­lights were and why they mat­ter more broad­ly. Read on for our conversation:

Q: What Kan­vas Bio plat­form updates are you most proud of?

A, Hao: We’ve already built the tools to under­stand how micro­bial cells inter­act with each oth­er and host cells. But now we’re able to lever­age this under­stand­ing to engi­neer com­plex, micro­bial ther­a­pies. Being able to visu­al­ize hun­dreds of micro­bial species in a sin­gle imag­ing round is a unique capa­bil­i­ty, as is being able to engi­neer a diverse ecosys­tem that resem­bles a healthy, human gut micro­bio­me in a sin­gle man­u­fac­tur­ing run. Because of these unique capa­bil­i­ties, our plat­form is posi­tioned to pro­vide a bet­ter mech­a­nis­tic under­stand­ing of micro­bio­me-relat­ed dis­eases, and improve effi­ca­cy for micro­bio­me-direct­ed therapies.

Q: What are Kan­vas Bio’s biggest sci­en­tif­ic achieve­ments this year?

A, Phil: Our R&D team has been work­ing dili­gent­ly to extend the Kan­vas Bio plat­form to accel­er­ate the pace of bio­log­i­cal dis­cov­ery for ther­a­peu­tic appli­ca­tions. This includes the abil­i­ty to rapid­ly screen can­di­date live bio­ther­a­peu­tics prod­ucts (LBPs) and mea­sure gene expres­sion. Up until now, no one in spa­tial biol­o­gy has been able to com­mer­cial­ly pair the micro­bial ecol­o­gy aspect with host gene expres­sion, so we’re real­ly excit­ed to debut this capability.

Host-microbe biol­o­gy is essen­tial­ly a black box. Sequenc­ing allows us to under­stand asso­ci­a­tions between microbes and their envi­ron­ment, but there hasn’t been a way to eas­i­ly probe mech­a­nisms. Kan­vas Bio is solv­ing this prob­lem by being able to deter­mine the place and func­tion of each microbe. By unlock­ing an under­stand­ing of the func­tion and posi­tion of microbes rel­a­tive to host cells and each oth­er, we can offer new ther­a­peu­tic modal­i­ties to alter dis­ease course in numer­ous indi­ca­tions. For exam­ple, if we can deter­mine that cer­tain bac­te­ria are phys­i­cal­ly asso­ci­at­ed with par­tic­u­lar immune cells rel­e­vant to a dis­ease, we can use those bac­te­ria as a ther­a­peu­tic tar­get or vehicle.

Q: How is Kan­vas Bio lever­ag­ing machine learn­ing and AI?

A, Hao: We use AI exten­sive­ly in image pro­cess­ing and spec­tral analy­sis. We have the largest spa­tial and func­tion­al host-micro­bio­me dataset in the world, so we’re eager to con­tin­ue learn­ing from this data, and use the knowl­edge to inform our approach to design­ing micro­bial ther­a­pies. Right now, we’re adapt­ing exist­ing AI frame­works for our use cas­es – one of which is seg­ment­ing bac­te­ria. Machine learn­ing mod­els are only as good as the data they’re trained on, so we have to do a fair amount of tweak­ing upfront as the data we work with (com­plex micro­bial com­mu­ni­ties) is very unique. In 2024, we plan to spend more time with large lan­guage mod­els (LLMs) to see how they can ben­e­fit the micro­bio­me field.

Q: Look­ing ahead to 2024, what R&D is Kan­vas Bio prioritizing?

A, Phil: We’re going to fur­ther devel­op the Kan­vas Bio plat­form to mea­sure micro­bial gene expres­sion, in addi­tion to host gene expres­sion and micro­bial tax­o­nom­ic iden­ti­ty. This will allow us to under­stand how dif­fer­ent microbes change in response to dif­fer­ent dis­eases and in the prox­im­i­ty to dif­fer­ent cell types. We’ll also begin man­u­fac­tur­ing KAN-001, our lead drug can­di­date in our Immuno-oncol­o­gy Program. 

Addi­tion­al­ly, in the com­ing year we’re extreme­ly excit­ed to lever­age the Kan­vas Bio plat­form to opti­mize LBP design and man­u­fac­tur­ing. With our East and West coast team mem­bers work­ing syn­er­gis­ti­cal­ly, we’ve already come up with sev­er­al ways our imag­ing tech­nol­o­gy can be extend­ed to improve iso­la­tion and manufacturing. 

Q: How did the Fed­er­a­tion Bio acqui­si­tion ben­e­fit Kan­vas Bio’s technology? 

A, Hao: Acquir­ing key Fed­er­a­tion Bio assets moved our time­line for ther­a­peu­tic devel­op­ment ahead by sev­er­al years. We gained access to a strong IP port­fo­lio, a star team for com­plex micro­bio­me design and man­u­fac­tur­ing, and a diverse micro­bial strain iso­late library. The strain library, in par­tic­u­lar, has real­ly boost­ed our exist­ing tech, because it’s high­ly diverse AND we know exact­ly what each strain is. This allows us to test dri­ve and stress test key com­po­nents of our plat­form tech­nol­o­gy in a well-con­trolled manner. 

Q: What biotech trends are top of mind for Kan­vas Bio right now?

A, Phil: Gen­er­at­ing, stor­ing, pro­cess­ing and ana­lyz­ing large amounts of bio­log­i­cal data excites us. Recur­sion Phar­ma­ceu­ti­cals set a great prece­dent for lever­ag­ing low dimen­sion­al data to gen­er­ate nov­el insights (and using those insights to cre­ate won­der­ful drugs from their imag­ing plat­form). Kan­vas Bio is gen­er­at­ing hun­dreds of giga­bytes of high dimen­sion­al data every day, and we’re using these data to answer spe­cif­ic ques­tions. We also have an abun­dance of hid­den” data that we’re not yet gen­er­at­ing insights from, so we’ll be con­tin­u­ing to expand our data­base of high-dimen­sion­al images and inte­grat­ing LLMs to query it. 

Q: This is Kan­vas Bio’s third Sci­ence Day, how does it com­pare to the inau­gur­al event?

A, Phil: We’ve come SO far in just three years. 2023 Sci­ence Day was a lot more pol­ished, and we had so many more tech­nol­o­gy and man­u­fac­tur­ing updates to share! It’s thrilling to grow so rapid­ly, but it’s even more spe­cial to be a part of the Kan­vas Bio team. We’re deeply pas­sion­ate sci­en­tists and that pas­sion is dri­ving our sci­ence for­ward in sub­stan­tial ways. Every day, we think about the mea­sur­able impact our work will have on human health and it moti­vates us to no end.