Great models demand great data

Rhys Grinter at University of Melborne discusses how bacterial pathogens encode proteins to extract iron, using Twist Clonal Genes to screen new binding proteins.

Felix Lansing at SeamlessTx explores how data-driven approach accelerates the generation of high-performance recombinases, thereby expanding the therapeutic potential of gene editing.

Fong Tian Wong at A*STAR IMCB explores how synthetic biology and molecular engineering can activate silent biosynthetic gene clusters in actinobacteria, streamlining metabolite production and enabling the discovery of novel compounds.

Ryan Emerson at A-Alpha Bio explores large-scale measurements of protein-protein interactions to build the data foundation for next generation discovery.

Using Gene Fragments and Pooled DNA, A-Alpha Bio generates massive, high-quality wet-lab datasets to train its AlphaBind AI model, enabling the rapid and simultaneous optimization of complex therapeutic antibody properties in a single, data-driven step.

Seeking to screen a large number of computationally generated antibodies, the David Baker Lab turned to Twist’s multiplex gene fragments.

Using Gene Fragments, researchers integrated wet-lab results to refine Machine Learning models, accelerating antibody optimization and ensuring accuracy in a closed-loop system.

Ryan Emerson at A-Alpha Bio explores large-scale measurements of protein-protein interactions to build the data foundation for next generation discovery.

Vladimir Gligorijevic at Genentech explores how combining AI-driven design with targeted lab testing accelerates the discovery of high-performance therapeutic antibodies, reducing experimental burden and improving candidate quality.

Twist and Cradle discuss how to build an iteration-focused lab environment to rapidly validate and improve machine learning models for biologics discovery.

Phytoform Labs used Twist’s synthesis solutions to streamline trait selection and optimize R&D resources.

Using Twist’s clonal genes and gene fragments, Resurrect Bio scaled the screening of thousands of AI-designed genes — building an efficient funnel for trait discovery, from in silico prediction to in planta validation.

Prime editing is one of the newest CRISPR technologies. It is highly promising as it can install small and precise mutations with a unique reverse transcriptase dependent mechanism that eliminates the need for donor DNA and double-strand breaks. Due to its precise editing capabilities, it has potential applications in CRISPR therapeutics for treating genetic disease.

Ci Chu at Xaira explores new approaches to generating large-scale, single-cell perturbation datasets that overcome key challenges in throughput, variability and batch effects.

To address the bottlenecks in genomic modeling and design, Brian Hie discussed his work on the genomic foundation model Evo2.