datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
Proteinscasp14-casp15-cameo-test-proteinsProteinSelfies10 million random examples from Uniref50 representative sequences (October 2023) and computed selfies strings. The strings are stored as input ids from a custom selfies tokenizer. A BERT tokenizer with this vocabulary has been uploaded to this dataset under the files.
You can access the tokenizer like this:
import os
from huggingface_hub import hf_hub_download
from transformers import AutoTokenizer
repo_path = 'Synthyra/ProteinSelfies'
local_path = 'ProteinSelfies'
files =… See the full description on the dataset page: https://huggingface.co/datasets/Synthyra/ProteinSelfies.protein-secondary-structure-netsurfp
NetSurfP-3.0 Secondary-Structure Splits
This dataset repo contains NetSurfP-derived protein secondary-structure labels
converted for Protein-I-JEPA probe training and evaluation.
Source page: https://services.healthtech.dtu.dk/services/NetSurfP-3.0/5-Dataset.php
Profile: hhblits
Labels are Q3 per-residue labels:
H: helix
E: beta strand
C: coil/other
.: ignored residue for loss and accuracy
Splits
Split
Rows
JSONL
TSV
train
10348
train.jsonl
tsv/train.tsv… See the full description on the dataset page: https://huggingface.co/datasets/lamm-mit/protein-secondary-structure-netsurfp.protein_structure_NER_independent_val_set
Overview
This data was used to evaluate the two models below to decide whether convergence was reached.
https://huggingface.co/PDBEurope/BiomedNLP-PubMedBERT-ProteinStructure-NER-v2.1
https://huggingface.co/PDBEurope/BiomedNLP-PubMedBERT-ProteinStructure-NER-v3.1
There are 20 different entity types in this dataset:
"bond_interaction", "chemical", "complex_assembly", "evidence", "experimental_method", "gene",
"mutant", "oligomeric_state", "protein", "protein_state", "protein_type"… See the full description on the dataset page: https://huggingface.co/datasets/PDBEurope/protein_structure_NER_independent_val_set.protein_structure_NER_independent_val_set
Overview
This data was used to evaluate the two models below to decide whether convergence was reached.
https://huggingface.co/PDBEurope/BiomedNLP-PubMedBERT-ProteinStructure-NER-v2.1
https://huggingface.co/PDBEurope/BiomedNLP-PubMedBERT-ProteinStructure-NER-v3.1
There are 20 different entity types in this dataset:
"bond_interaction", "chemical", "complex_assembly", "evidence", "experimental_method", "gene",
"mutant", "oligomeric_state", "protein", "protein_state", "protein_type"… See the full description on the dataset page: https://huggingface.co/datasets/mevol/protein_structure_NER_independent_val_set.protein_stability_single_mutation
Protein Data Stability - Single Mutation
This repository contains data on the change in protein stability with a single mutation.
Attribution of Data Sources
Primary Source: Tsuboyama, K., Dauparas, J., Chen, J. et al. Mega-scale experimental analysis of protein folding stability in biology and design. Nature 620, 434–444 (2023). Link to the paper
Dataset Link: Zenodo Record
As to where the dataset comes from in this broader work, the relevant dataset (#3) is shown in… See the full description on the dataset page: https://huggingface.co/datasets/Trelis/protein_stability_single_mutation.pyaptamer-proteins-shin2023Dataset-Structure_Class-ProteinShake
Description
Structural Class Prediction is a multi-class classification task to predict the correct structural class of given protein. This task is is built on the SCOP database.
Protein Format: SA sequence (PDB)
Splits
The dataset is from ProteinShake Building datasets and benchmarks for deep learning on protein structures. We use the splits based on 70% structure similarity, with the number of training, validation and test set shown below:
Train: 7990
Valid: 955
Test:… See the full description on the dataset page: https://huggingface.co/datasets/SaProtHub/Dataset-Structure_Class-ProteinShake.protein-structure-trust-benchmark
Protein-Structure Trust-Routing Benchmark (Boltz-2)
Leakage-controlled benchmarks for confidence-calibrated trust routing over a protein-structure
predictor: given a specialist model's confidence (Boltz-2 ipTM / pLDDT) for a target, decide whether to
trust the prediction or pay to verify it — and score that decision against experimentally-measured
correctness. Evaluation substrate for the report "When does an LLM trust a specialist model? A cost-aware
trust-routing audit"… See the full description on the dataset page: https://huggingface.co/datasets/jang1563/protein-structure-trust-benchmark.greenbeing-proteins
GreenBeing Proteins dataset
Proteins from UniProtKB (knowledge base), from select food crops and related species.
Amino acid sequences use IUPAC-IUB codes where letters A-Z map to amino acids.
Usage (due to different schema on splits):
load_dataset("monsoon-nlp/greenbeing-proteins", "pretraining", split="pretraining")
XML source from https://www.uniprot.org/help/downloads
CoLab notebook: https://colab.research.google.com/drive/1M6sO0Ws6i5z9VUXIXopiOqo1OkQ7K-1g?usp=sharing… See the full description on the dataset page: https://huggingface.co/datasets/monsoon-nlp/greenbeing-proteins.tda_for_proteinsDataset-Structural_Similarity-ProteinShake
Description
Structure Similarity Prediction predicts the (aligned) Local Distance Difference Test (LDDT) of the structures given an unaligned pair of proteins. Target values are computed after alignment with TM-align for all pairs of 1000 randomly sampled single-chain proteins.
Protein Format: SA sequence(PDB)
Splits
The dataset is from ProteinShake Building datasets and benchmarks for deep learning on protein structures. We use the splits based on 70% structure… See the full description on the dataset page: https://huggingface.co/datasets/SaProtHub/Dataset-Structural_Similarity-ProteinShake.aging_proteins
Description of the Dataset
This is (part of) the dataset used in
Prediction and characterization of human ageing-related proteins by using machine learning.
This can be used to train a binary sequence classifier using protein language models such as ESM-2.
Please also see the github for the paper for more information.
protein-sft-uniprot
protein-sft-uniprot
단백질 연구 특화 LLM 학습을 위한 SFT(Supervised Fine-Tuning) 데이터셋입니다.
UniProt 데이터베이스와 단백질 문헌에서 추출한 Q&A 형식의 대화 데이터로 구성되어 있습니다.
Dataset Summary
Total samples
1,551,711
Unique proteins
455,613
Format
JSONL (chat messages)
Size
462MB
Sources
Source
Samples
Description
UniProtQA
1,513,126
UniProt 데이터베이스에서 구조화된 단백질 정보 추출
Protein2Text-QA
38,585
단백질 관련 문헌 기반 심화 Q&A
Question Types
Type
Samples
Example… See the full description on the dataset page: https://huggingface.co/datasets/im-sangwoon/protein-sft-uniprot.gran-nanobody-proteins
GRAN Protein Structure Dataset
Dataset Description
This dataset contains protein graph data for training Graph Recurrent Attention Networks (GRAN) for protein sequence and structure generation.
Dataset Summary
Number of proteins: 2965
Average protein length: 121.0 residues
Unique amino acids: 22
Source: Nanobody protein structures
Created by: alexchilton
Date: 2025-05-04 16:01:38
Dataset Structure
Each protein entry contains:
sequence: Complete… See the full description on the dataset page: https://huggingface.co/datasets/alexchilton/gran-nanobody-proteins.protein_solubility_classprimate-proteinsAll primate proteins in UniProtKB (knowledge base) which have been reviewed (Swiss-Prot), including but not limited to humans.
Each row contains a gene name, species or subspecies, an amino acid sequence, and comments / annotations available in UniProt.
A gene name may match multiple entries / IDs on UniProt from different accessions.
Annotations may be empty, or may include information such as:
likely function, written in English
location inside of cell (e.g. "Subcellular locations:… See the full description on the dataset page: https://huggingface.co/datasets/monsoon-nlp/primate-proteins.Dataset-Binding_Site_Detection-ProteinShake
Description
Binding Site Detection predicts , predict whether a protein residue belongs to a small molecule binding cavity. Binding site residues are those within the binding pocket provided by PDBBind. Default metric is Matthew's Correlation.
Splits
Protein Format: SA sequence (PDB)
The dataset is from ProteinShake Building datasets and benchmarks for deep learning on protein structures. We use the splits based on 70% structure similarity, with the number of training… See the full description on the dataset page: https://huggingface.co/datasets/SaProtHub/Dataset-Binding_Site_Detection-ProteinShake.e_coli_proteins_bacteriaCDS-BART-Ecoli-proteins📊 E.coli protein dataset
This dataset comprises E.coli protein expression levels, categorized in low, medium and high.
Specifically, it contains 2308 low, 2067 medium and 1973 high expression data facilitating the study of protein production in bacterial cells.
The original dataset is from CodonBERT.
⁉️ Dataset Contents
Sequence: The mRNA sequence of the mRNA degradation
Expression: The three levels of expression data
🎯 Purpose
This dataset serves as a benchmark for fine-tuning models on a… See the full description on the dataset page: https://huggingface.co/datasets/mogam-ai/CDS-BART-Ecoli-proteins.protein-ss-datasetprimate-proteins
Dataset Transformation Summary
Original Dataset: monsoon-nlp/primate-proteinsTransformed Dataset: pkanithi/primate-proteins
Changes Made
Added messages Column
Added a new messages column in ChatML format
Each example now contains a conversation structure with system, user, and assistant messages
Data Filtering
Filtered out proteins with no annotation (annotation != None)
Ensures all examples have valid ground truth annotations
Dataset… See the full description on the dataset page: https://huggingface.co/datasets/pkanithi/primate-proteins.pdb_id_Q_and_len
Explore & contribute. Join a discovery effort
This dataset is part of a citizen science project proposing the Mosaic Q pattern:
a conserved structural trait conserved in protein structure, where amino
acids cluster by chemical family in groups of similar size and shape.
Use our interactive app to visualise any protein in this dataset and contribute your
observation to our collaborative repository:
👉 Open the Proteins Mosaic Q Space
Enter a PDB ID → inspect the 3D structure →… See the full description on the dataset page: https://huggingface.co/datasets/ProteinsMosaicQ/pdb_id_Q_and_len.protein_secondary_structure_from_PDBThis dataset contains 125,955 protein sequences, with protein PDB ID, length, the sequence (primary structure), as well as secondary structure as identified from experiment. The shortest protein is composed of only 11 amino acids, along with the longest one that features up to 19,350 amino acids. The standard deviation of the length is 855 amino acids.
The dataset further includes overall secondary sturctrure content, for all eight classes of secondary structure types.
The beta sheet content… See the full description on the dataset page: https://huggingface.co/datasets/lamm-mit/protein_secondary_structure_from_PDB.ProteinSpace-TheraHuman-mAbs
Dataset Description
ProteinSpace-TheraHuman-mAbs is a curated collection of therapeutic antibody sequences derived from the Thera-SAbDab database. The dataset contains 1,400 antibody chain sequences (700 heavy chains and 700 light chains) from 700 therapeutic antibodies, all of which are either genetically human or humanized whole monoclonal antibodies (mAbs).
Each sequence has been processed with ANARCI (Antibody Numbering and Receptor ClassIfication) to provide IMGT-numbered… See the full description on the dataset page: https://huggingface.co/datasets/melanierb/ProteinSpace-TheraHuman-mAbs.protein_structure_pathogenicity_dataset
Protein Structure Pathogenicity Dataset
Dataset Description
This dataset contains protein structures and metadata for benign and pathogenic missense variants, designed for training machine learning models to predict variant pathogenicity using protein structural information.
Dataset Summary
The dataset includes:
Protein 3D structures predicted via ESMFold
Benign and pathogenic variants derived from the ProteinGym benchmark
Structural and sequence… See the full description on the dataset page: https://huggingface.co/datasets/CHGGM-Aachen/protein_structure_pathogenicity_dataset.protein-secondary-structure-nppe2
Protein Secondary Structure Prediction Dataset (NPPE-2)
Dataset Description
This dataset is used for the Protein Secondary Structure Prediction competition from the Deep Learning and Generative AI course at IIT Madras.
Task
Predict the secondary structure of proteins at the residue level using:
Q8 (sst8): Eight-state secondary structure classification (DSSP notation)
Q3 (sst3): Three-state secondary structure classification
Evaluation Metric… See the full description on the dataset page: https://huggingface.co/datasets/lamm-mit/protein-secondary-structure-nppe2.protein-secondary-structure-predictswissprot-proteins
Uniprot SwissProt v. 2024_01
List of protein sequences and selected protein-level annotations for SwissProt v. 2024_01.
References:
UniProt: the Universal Protein Knowledgebase in 2025. The UniProt Consortium. https://doi.org/10.1093/nar/gkae1010