Protein Diversity

Evaluate amino acid diversity in a protein sequence

This constraint is open source. Any third-party models, product names, or trademarks referenced are the property of their respective owners, and Proto is not affiliated with them.

---

Source

proto-bio/proto-language/proto_language/constraint/protein_quality/protein_diversity_constraint.py

View source

Evaluate amino acid diversity in protein sequences.

This constraint function measures the diversity of amino acid types present in protein sequences. It calculates diversity as the fraction of the 20 standard amino acids that appear in the sequence, and penalizes sequences that fall below a minimum diversity threshold. The penalty scales linearly with the deficit below the minimum diversity threshold.

API Reference

ConfigProteinDiversityConfig Source

Configuration for protein diversity constraint.

This class defines configuration parameters for evaluating amino acid diversity in protein sequences. The constraint measures how many different amino acid types are present in the sequence and penalizes sequences with insufficient diversity, which may indicate poor protein quality, repetitive sequences, or non-functional proteins.

A diversity score of 1.0 means all 20 standard amino acids are present. The minimum for a non-empty sequence is 0.05 (1/20), reached by a homopolymer (only one amino acid type); a score of 0.0 is unreachable since empty sequences are rejected.

min_diversity · number · default: "0.7"

Minimum acceptable amino acid diversity. Calculated as (unique amino acids) / 20.

ReturnsConstraintOutput

One result per sequence. A score of 0.0 indicates sufficient diversity (diversity at or above threshold) and higher values indicate insufficient amino acid diversity. Scores scale linearly with the deficit below the threshold (e.g., if min_diversity is 0.5 and actual diversity is 0.25, the score is 0.5), capped at 1.0. metadata carries:

• aa_diversity_score: Float diversity score (0.0-1.0) calculated as (unique amino acids) / 20
• unique_amino_acid_count: Integer count of unique amino acid types present in the sequence (0-20)
• unique_amino_acids: Sorted list of amino acid characters present in the sequence

Usage

Evaluating protein diversity:

python

>>> from proto_language.core import Sequence, SequenceType
>>> config = ProteinDiversityConfig(min_diversity=0.5)
>>> seq = Sequence("MVLSPADKTNVKAAWGKVGAHAGEYGAEALERMFLSF", "protein")
>>> results = protein_diversity_constraint([(seq,)], config)
>>> print(results[0].score)  # 0.0 (diversity 0.85 >= 0.5)
>>> print(results[0].metadata["aa_diversity_score"])  # 0.85
>>> print(results[0].metadata["unique_amino_acid_count"])  # 17
>>> print(results[0].metadata["unique_amino_acids"])  # ['A', 'D', 'E', 'F', ...]

Metadata

Property	Value
Key	protein-diversity
Function	protein_diversity_constraint
Category	protein_quality
Mode	discrete
Uses GPU	False
Supported Types	protein