Uri Laserson

I shepherded Patch Bio's integration into Ginkgo, including transfer of datasets and protocols to various groups at Ginkgo, delivery of IP and sales materials for BD purposes, adoption of Patch Bio data management processes for pooled workflows, and wind-down of our New York site. I also worked to define the research directions for the AI group.

I co-founded Patch Bio, a techbio company that combined high-throughput sequencing-based pooled assays with machine learning to discover synthetic regulatory elements (promoters, UTRs) for nucleic acid therapeutics (AAV, mRNA). I defined and led the R&D strategy, managing a team of up to 15 wet-lab and computational scientists. My work spanned product strategy, experiment design and troubleshooting, data analysis, machine learning model development, and sometimes being the IT guy or furniture assembler. I played an active role in business development and operations, pitching to investors and strategic partners, handling IP matters, and hiring. We raised $11.5M in seed funding and secured a major pharma deal. Patch Bio was acquired by Ginkgo Bioworks in March 2024.

I helped to define the initial scientific direction and hire the initial computational team.

I built and ran a genetics/immunology research lab with a ~$500k annual budget, overseeing a team of up to five trainee scientists. I defined the lab's strategic direction, planned and analyzed experiments, developed collaborations, raised funding, advised graduate students and postdocs, and taught classes in the graduate program. I was awarded a $3M NIH R01 grant (declined) to develop computational methods for high-throughput immunological assays (PhIP-seq). I also received a $200k grant from the Chan Zuckerberg Initiative to prototype distributed computing engines for single-cell RNA-seq analysis. I published six papers, processed around 1,000 patient samples (generating over 600 GB of sequencing data), presented at multiple conferences, and led a consortium working group focused on data standardization for immune repertoire data.

I led the development of impyla, an open-source Python client for the Impala query engine, and implemented an LLVM compiler target for compiled Python UDFs, along with a distributed dataframe abstraction. As a committer to BigDataGenomics and ADAM, I worked on large-scale genomics computation on Spark and deployed genomics solutions on Hadoop at several life sciences organizations. In addition to technical work, I provided expert guidance in scalable genomics, healthcare, and machine learning during client consulting engagements and sales support. I coauthored research articles, the O’Reilly book "Advanced Analytics on Spark", and wrote multiple technical blog posts, including the most-read post on Cloudera’s blog at the time.

AbVitro was founded partly based on my graduate research.

I spearheaded the application of next-generation sequencing (NGS) technology in immunology, increasing the number of known antibody sequences by 100-fold. I designed and implemented statistical analysis methods for high-throughput autoantigen discovery, published in Nature Biotech. I also devised and tested various experimental approaches for capturing antibodies at scale, using techniques such as emulsions, microfluidics, and in situ biochemistry. I also developed a software pipeline to manage large antibody sequence datasets, including alignment, clustering, and database functionalities. Throughout my time, I initiated and managed multiple research projects and collaborations with leading labs, including those of Steve Elledge (Harvard), Dennis Burton (Scripps), and Daphne Koller (Stanford).

I defined the early technological direction of Good Start Genetics, developed a comprehensive R&D plan, analyzed competing technologies, designed initial budgets, and recruited the first technical hires. I also pitched to VCs and angel investors, successfully raising $1M in initial capital and leading the company through an $18M Series A round. In total, GSG raised $60M before being acquired by Invitae in 2017.