Protein structure-based protein language models

A non-transitory computer readable storage medium impressed with computer program instructions to score nucleotide variants as benign or pathogenic, the instructions, when executed on a processor, implement a method comprising:

• accessing a protein that has respective amino acids at respective positions;
• specifying a particular amino acid at a particular position in the protein as a gap amino acid, and specifying remaining amino acids at remaining positions in the protein as non-gap amino acids;
• generating a gapped spatial representation of the protein that
• includes spatial configurations of the non-gap amino acids, and
• excludes a spatial configuration of the gap amino acid;
• and
• using a neural network pathogenicity predictor to score a nucleotide variant as benign or using a based at least on
• the gapped spatial representation, and
• a representation of an alternate amino acid created by the nucleotide variant at the particular position.

A system including one or more processors coupled to memory, the memory loaded with computer instructions to score nucleotide variants as benign or pathogenic, the computer instructions, when executed on the one or more processors, implement actions comprising:

• accessing a protein that has respective amino acids at respective positions;
• specifying a particular amino acid at a particular position in the protein as a gap amino acid, and specifying remaining amino acids at remaining positions in the protein as non-gap amino acids;
• generating a gapped spatial representation of the protein that
• includes spatial configurations of the non-gap amino acids, and
• excludes a spatial configuration of the gap amino acid;
• and
• using a neural network pathogenicity predictor to score a nucleotide variant as benign or using a based at least on
• the gapped spatial representation, and
• a representation of an alternate amino acid created by the nucleotide variant at the particular position.