MOTIVATION LETTER Open Cell’s model of short-term, flexible lab rentals for early-stage life science ventures is the missing piece in my current validation pipeline. My venture predicts drug resistance mutations from protein sequence alone, using an ESM-2 protein language model with delta-embeddings combined with ECFP4 drug fingerprints and a Random Forest classifier. The model achieves an AUROC of 0.804 plus or minus 0.025 on the Platinum benchmark with 553 mutations under protein-grouped cross-validation, and 0.634 on SKEMPI 2.0. It covers 100 percent of mutations in its test space, compared to roughly 18 percent for structure-limited tools, and it beats published state of the art, with mCSM-lig at approximately 0.70 AUROC. The venture is pre-seed, proof-of-concept validated, and not yet incorporated. I am a pharmacist turned machine learning engineer and the sole author of the codebase and the research. The immediate technical roadmap is to fine-tune ESM-2 on the SKEMPI 3K mutation set to reach an AUROC of at least 0.70 on that benchmark, then move into a pilot with Servier in Suresnes. The scientific bottleneck is wet lab validation, not compute. Predicted resistance mutations need to be tested in vitro to confirm that the model’s outputs correspond to real biological outcomes. That requires bench space, cell culture access, and shared lab services, which I do not currently have. Open Cell’s mission to support early-stage life science innovation through affordable, flexible lab space directly addresses this gap. The venture is computational, but it is not purely in silico. The commercial path depends on generating experimental evidence that the predictions hold. Open Cell’s shared services and short-term rental model fit a solo founder who needs intermittent bench access rather than a full institutional lease. The selection criteria for Open Cell include early-stage biotech focus, need for wet lab space, commercial application potential, and team capability. I meet each of these. The venture is early stage, pre-revenue, and pre-incorporation. It has a validated proof of concept with quantitative benchmarks. The commercial application is clear: pharmaceutical companies and biotech firms need to know which resistance mutations will emerge for a given drug candidate, and current tools miss most of the sequence space. The team capability is demonstrated by the fact that a single founder built and validated the model, published the benchmarks, and is now pursuing named partnerships with Servier, Paris-Saclay I2BC, Institut Pasteur, and Sanofi in Gentilly. I am applying to Open Cell because the venture needs what Open Cell provides: affordable lab space to run validation experiments on predicted resistance mutations, and a community of early-stage life science founders. The support pipeline already includes SEMIA and Quest for Health meetings in progress, WILCO One BioTech in October 2026, and submitted applications to IncubAlliance and AI House. Open Cell is the wet lab complement to those computational and business development resources. The venture’s focus on drug resistance mutations addresses a critical unmet need in drug discovery. Resistance is the reason most cancer therapies and antimicrobials fail over time. Predicting resistance mutations before they emerge in the clinic allows drug developers to design combination therapies, plan for resistance monitoring, and make better go or no-go decisions. The model does this from sequence alone, which means it works for targets where no crystal structure exists, which is the majority of the proteome. That is the commercial wedge, and Open Cell is the right environment to generate the validation data that turns the wedge into a product. APPLICATION SUMMARY Venture name: The venture, pre-seed AI and biotech startup, not yet incorporated. Founder: Eniola Olutogun, pharmacist and machine learning engineer, sole author. Technology: ESM-2 protein language model delta-embeddings plus ECFP4 drug fingerprints, Random Forest classifier. Predicts drug resistance mutations from protein sequence alone. No crystal structure required. Validation status: Proof of concept validated. AUROC 0.804 plus or minus 0.025 on Platinum benchmark, 553 mutations, protein-grouped cross-validation. AUROC 0.634 on SKEMPI 2.0. 100 percent mutation coverage versus approximately 18 percent for structure-limited tools. Beats published SOTA, mCSM-lig at approximately 0.70 AUROC. Wet lab need: Validation of predicted resistance mutations in vitro. Requires bench space, cell culture, shared lab services. Commercial path: Fine-tune ESM-2 on SKEMPI 3K mutations to AUROC at least 0.70, then Servier pilot, then annual recurring revenue from pharma licensing. Named partners: Servier, Suresnes. Paris-Saclay I2BC. Institut Pasteur. Sanofi, Gentilly. Support pipeline: SEMIA and Quest for Health meeting in progress. WILCO One BioTech October 2026. IncubAlliance and AI House applications submitted. EIC Accelerator and BPI i-Lab as future targets. Why Open Cell: Short-term flexible lab rentals and shared services match the venture’s need for intermittent wet lab access. Alignment with Open Cell’s mission to support early-stage life science innovation. Commercial application potential is clear. Team capability demonstrated by validated proof of concept and named partnerships. EMAIL DRAFT Subject: Open Cell application, computational biotech venture seeking wet lab validation space Dear Open Cell team, My venture predicts drug resistance mutations from protein sequence alone, using an ESM-2 protein language model with delta-embeddings and ECFP4 drug fingerprints. The model achieves an AUROC of 0.804 on the Platinum benchmark and covers 100 percent of mutations versus roughly 18 percent for structure-limited tools. It is pre-seed, proof-of-concept validated, and I am the sole founder, a pharmacist turned machine learning engineer. The venture is computational, but the next validation step is wet lab work. Predicted resistance mutations need in vitro confirmation before pharmaceutical partners will take the model seriously. I have a pilot discussion in progress with Servier in Suresnes, and named relationships with Paris-Saclay I2BC, Institut Pasteur, and Sanofi in Gentilly. What I lack is bench space and shared lab services. Open Cell’s short-term flexible lab rentals and shared services fit exactly what the venture needs. I do not require a full institutional lease. I need intermittent access to cell culture and bench space to run validation experiments on predicted mutations. Open Cell’s model of supporting early-stage life science innovation is the right environment for that work. I am applying for a place in the Open Cell incubator. The venture has a validated proof of concept, a clear commercial path through pharma licensing, and a concrete need for the facilities Open Cell provides. I would welcome the chance to discuss how the venture fits your community. Best regards, Eniola Olutogun CHECKLIST - [ ] Confirm Open Cell application deadline and submission portal from programme website - [ ] Verify whether Open Cell requires a formal application form in addition to email - [ ] Confirm whether Open Cell accepts pre-incorporation ventures - [ ] Prepare one-page technical summary of the model and benchmarks for Open Cell review - [ ] Prepare summary of Servier pilot discussion status and named partner relationships - [ ] Confirm wet lab validation experiment scope and estimated bench time required - [ ] Insert personal background detail, including pharmacy degree institution and ML engineering experience - [ ] Confirm whether Open Cell requires a pitch deck or financial projections - [ ] Verify Open Cell rental pricing and shared services list before committing - [ ] Send email draft to Open Cell contact, or adapt to application form if required EDITOR NOTES - Eligibility risk: Open Cell may require ventures to be incorporated or to have a team of more than one founder. The venture is pre-incorporation and solo founder. Verify before submitting. - Facts to verify: The AUROC numbers and mutation coverage figures are from the profile and should be checked against the actual benchmark results before they appear in any formal application. The Servier pilot is described as in progress, not confirmed. Confirm the current status. - Gaps for applicant: The profile does not include the founder’s pharmacy degree institution, ML engineering training background, or any prior startup experience. Insert these details where the application asks for team background. - The email draft is written for a cold outreach format. If Open Cell’s website provides a specific application form or asks for a formal proposal, adapt the content into that format rather than sending the email as is. - The venture is computational but requires wet lab validation. Do not overstate the wet lab component. Open Cell is being asked for intermittent bench access, not a full laboratory program. Keep that framing consistent in any verbal or written follow-up.