pydssp
A simplified implementation of DSSP algorithm for PyTorch and NumPy
Decision gist · record as of 2026-08-14
Yes, if you need secondary structure assignment for protein backbones and can work with a simplified DSSP that omits β-bulge annotation. The 97% agreement with original DSSP is suitable for most applications. Install friction is low and the package is actively maintained. No security vulnerabilities are known. Choose this if you need PyTorch differentiability or batch processing; otherwise, consider the original DSSP or STRIDE for full feature parity.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires PyTorch and NumPy installed; coordinates must be shaped as (batch, length, 4 or 5, 3) where atoms are N, CA, C, O backbone atoms (and optionally H).
- Low friction: pure Python wheel with four runtime dependencies (numpy, torch, einops, tqdm).
- Last release was 2025-02-24 and repository is active, though maintenance status is aging with no release in the past 536 days.
License · maintenance · safety
MIT (permissive) — MIT license (permissive) places no restrictions on use, modification, or distribution in commercial or private projects.
last release 2025-02-24 (536 days) · last repo commit 2025-06-09 · 104 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 99,424 downloads/mo, #13,023 on PyPI
Alternatives
Verify before relying
pip install pydssp
import torch
import pydssp
coord = torch.randn([10, 100, 4, 3]) # batch, length, atoms (N,CA,C,O), xyz
dssp = pydssp.assign(coord, out_type='c3') # returns ['-', 'H', 'E', ...]- Whether the 97% agreement with original DSSP is sufficient for your specific downstream application.
- Performance characteristics on large protein structures or batches beyond typical use.
- Whether simplified parameters (hydrogen placement, no β-bulge support) affect your research domain.
What it is and what it does
PyDSSP is a Python implementation of the DSSP (Dictionary of Secondary Structure of Protein) algorithm that classifies each residue in a protein backbone as alpha-helix (H), beta-strand (E), or loop (-). It takes 3D atomic coordinates as input and outputs secondary structure assignments. The package is not a complete reimplementation of the original 1983 DSSP algorithm—it simplifies some parts, such as omitting β-bulge annotation and using slightly different hydrogen-placement parameters—but achieves over 97% agreement with the original on average.
The main innovation is support for both NumPy arrays and PyTorch tensors, with the latter enabling differentiable computation. PyDSSP can output hydrogen-bond matrices as continuous values in [0, 1] rather than binary assignments, making it suitable for machine learning pipelines where gradients flow through structure prediction. It includes a command-line tool for batch processing PDB files and a Python API for programmatic use.
Use it for
- Assign secondary structure to protein backbones in molecular dynamics simulations or structure prediction pipelines.
- Generate differentiable hydrogen-bond matrices for neural network training on protein structure tasks.
- Batch-process multiple PDB files to extract C3 secondary structure annotations via the command-line tool.
- Compute continuous hydrogen-bond scores for structure validation or quality assessment workflows.
- Integrate structure assignment into deep learning models where gradients must flow through secondary structure computation.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you need secondary structure assignment for protein backbones and can work with a simplified DSSP that omits β-bulge annotation.
The 97% agreement with original DSSP is suitable for most applications. Install friction is low and the package is actively maintained. No security vulnerabilities are known. Choose this if you need PyTorch differentiability or batch processing; otherwise, consider the original DSSP or STRIDE for full feature parity.
Install
pydssp on PyPI
Before you install
Low friction: pure Python wheel with four runtime dependencies (numpy, torch, einops, tqdm). Last release was 2025-02-24 and repository is active, though maintenance status is aging with no release in the past 536 days.
Requires PyTorch and NumPy installed; coordinates must be shaped as (batch, length, 4 or 5, 3) where atoms are N, CA, C, O backbone atoms (and optionally H).
License in practice
MIT license (permissive) places no restrictions on use, modification, or distribution in commercial or private projects.
Quickstart
pip install pydssp
import torch
import pydssp
coord = torch.randn([10, 100, 4, 3]) # batch, length, atoms (N,CA,C,O), xyz
dssp = pydssp.assign(coord, out_type='c3') # returns ['-', 'H', 'E', ...]Verify before relying
- Whether the 97% agreement with original DSSP is sufficient for your specific downstream application.
- Performance characteristics on large protein structures or batches beyond typical use.
- Whether simplified parameters (hydrogen placement, no β-bulge support) affect your research domain.
Package facts
| License | MIT permissive |
| Python support | Not specified |
| Install friction | Low. Pure-Python wheel |
| Runtime dependencies | 4 packagesnumpytorcheinopstqdm |
| Maintenance | Aging 536 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 99,424 / month, #13,023 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | License :: OSI Approved :: MIT LicenseProgramming Language :: Python :: 3.8 |
Evidence: pydssp-0.9.1-py3-none-any.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “protein secondary structure prediction”
- pydsspPyDSSP assigns protein secondary structure (alpha-helix, beta-strand,…
- mdtrajMDTraj reads, writes, and analyzes molecular dynamics trajectories in…
- fair-esmProvides pre-trained transformer protein language models (ESM-2,…
Give your agent the search over MCP, or paste the wish link into any chat.
More Artificial Intelligence packages
LiteLLM provides a unified Python interface to call 100+ LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Azure, and others) using OpenAI-compatible API format, available as both a Python SDK and a self-hosted AI Gateway proxy server.
Install it if you need to work with multiple LLM providers or want to centralize LLM routing in your organization.
Client library and CLI tool for downloading, uploading, and managing models, datasets, and repositories on the Hugging Face Hub platform.
Install it if you work with Hugging Face Hub models or datasets.
LangChain provides a framework for building agents and LLM-powered applications by composing language models, tools, and memory through a unified API that abstracts over multiple model providers.
hf-xet provides chunk-based deduplication and efficient file transfer for the Hugging Face Hub, enabling faster uploads and downloads of large files with local disk caching.
Tokenizers converts raw text into token sequences for NLP models, with support for training custom vocabularies and using pre-built tokenizers (BPE, WordPiece) optimized for speed via Rust.
Transformers provides a unified framework for loading, fine-tuning, and running state-of-the-art pretrained models across text, vision, audio, video, and multimodal tasks using PyTorch, JAX, or TensorFlow.
Install it if you need to run or train any transformer-based model for NLP, vision, audio, or multimodal tasks.
See also mdtraj · pdb2pqr · tmtools · ViennaRNA · fair-esm · propka · biotite · mhfp · mmcif-pdbx