Fig. 1: A deep learning approach for protease substrate design. | Nature Communications

Fig. 1: A deep learning approach for protease substrate design.

From: Deep learning guided design of protease substrates

Fig. 1: A deep learning approach for protease substrate design.The alternative text for this image may have been generated using AI.

A For a given peptide, the CleaveNet Predictor predicts cleavage scores across 18 matrix metalloproteinases (MMPs). B The CleaveNet Generator can generate substrates unconditionally (left) or conditionally, guided by a desired protease cleavage profile (right). C To design candidate substrates for a target protease, peptide sequences are generated by the CleaveNet Generator and prioritized by the cleavage scores from the CleaveNet Predictor. D An in vitro screen of 95 fluorogenic substrates against 12 recombinant MMPs was performed to validate the cleavage profiles of CleaveNet-designed substrates.

Back to article page