SPEAKING THE LANGUAGE OF PROTEINS
Our proprietary Foundation Model will drastically improve the success ratio of your leads in preclinical trials.
Our Technology
Our Core Technology is based in Protein Large Language Models (pLLMs) trained on 500 million unique protein-protein interactions and fine-tuned on 7000 antibody-antigen interactions for antibody-specific tasks, embedding a holistic set of protein information.
MALI™, Our Platform
Our core product is MALI™, modular software platform with AI solutions for therapeutic antibody design and optimization tasks.
The inputs are amino acid sequences of therapeutic targets, and the outputs are amino acid sequences of novel antibodies in any kind of antibody format against the therapeutic target of interest, or antibody mutants with improved biochemical properties.
Why is MALI™ Unique?
Due to the nature of our algorithm, in Silica Corpora we design successful candidates using a holistic approach, considering the system in their entirety rather than just focusing on a specific property. As a result, MALI is capable to take into account all set of parameters needed to create a successful candidate and modify them at the same time.
In addition, our models are specifically designed to be fine-tuned with project-specific data, and allows MALI to find the right answer in every project using only a very small data set, meaning that large amounts of data are not needed.
