Machine learning agents for Lobster AI (feature selection, survival analysis, interpretability)
Project description
lobster-ml
Machine learning and deep learning for biological data analysis and framework export.
Installation
# Basic installation
pip install lobster-ml
# With deep learning dependencies (scVI, PyTorch)
pip install lobster-ml[ml]
Agents
| Agent | Description |
|---|---|
machine_learning_expert |
ML specialist for biological data. Feature engineering, data splitting, framework export, and deep learning embeddings. |
Services
| Service | Purpose |
|---|---|
| MLPreparationService | Feature selection, scaling, and train/test/validation splitting |
| MLTranscriptomicsServiceALPHA | Transcriptomics-specific ML workflows (ALPHA) |
| MLProteomicsServiceALPHA | Proteomics-specific ML workflows (ALPHA) |
| scVIEmbeddingService | Deep learning embeddings using scVI for single-cell data |
Features
ML Readiness Assessment
- Evaluate biological datasets for machine learning suitability
- Check sample size, class balance, and feature quality
- Identify potential data leakage and batch effects
- Recommend preprocessing steps before ML pipeline
Feature Engineering
- Highly variable gene selection for dimensionality reduction
- PCA-based feature extraction with variance thresholds
- Marker gene features from differential expression
- Z-score normalization and scaling
Data Splitting
- Stratified train/test/validation splits
- Configurable split ratios (default: 70/15/15)
- Class balance preservation across splits
- Batch-aware splitting to prevent data leakage
Framework Export
- NumPy arrays for scikit-learn workflows
- CSV export for general ML frameworks
- PyTorch tensor datasets with DataLoader support
- TensorFlow NPZ format for Keras models
Deep Learning Embeddings
- scVI integration for variational autoencoder embeddings
- Latent space visualization and clustering
- Transfer learning from pre-trained models
- GPU acceleration when available
Requirements
- Python 3.12+
- lobster-ai >= 1.0.0
- Optional: torch, scvi-tools (for deep learning features)
Tier Requirement
This is a premium agent. Access is controlled at runtime via Lobster AI's tier system.
Documentation
Full documentation: docs.omics-os.com/docs/agents/ml
License
MIT
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file lobster_ml-1.1.418.tar.gz.
File metadata
- Download URL: lobster_ml-1.1.418.tar.gz
- Upload date:
- Size: 131.0 kB
- Tags: Source
- Uploaded using Trusted Publishing? Yes
- Uploaded via: twine/6.1.0 CPython/3.13.12
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
bcccb6f9f5fe8ad93cc4ce20deead9cf8103f4740d373672a6ec1df15dfd8e44
|
|
| MD5 |
e3fc50a3e9adc842023e37aee2a824e5
|
|
| BLAKE2b-256 |
f8b09b6107301c60f7c791f96ba148149e683bd4255ce1f07ccd81c10292ea45
|
Provenance
The following attestation bundles were made for lobster_ml-1.1.418.tar.gz:
Publisher:
publish-packages.yml on the-omics-os/lobster
-
Statement:
-
Statement type:
https://in-toto.io/Statement/v1 -
Predicate type:
https://docs.pypi.org/attestations/publish/v1 -
Subject name:
lobster_ml-1.1.418.tar.gz -
Subject digest:
bcccb6f9f5fe8ad93cc4ce20deead9cf8103f4740d373672a6ec1df15dfd8e44 - Sigstore transparency entry: 1429791942
- Sigstore integration time:
-
Permalink:
the-omics-os/lobster@2201d19fcf75f9b405154412a187b58a70ed66b6 -
Branch / Tag:
refs/tags/v1.1.418 - Owner: https://github.com/the-omics-os
-
Access:
public
-
Token Issuer:
https://token.actions.githubusercontent.com -
Runner Environment:
github-hosted -
Publication workflow:
publish-packages.yml@2201d19fcf75f9b405154412a187b58a70ed66b6 -
Trigger Event:
push
-
Statement type:
File details
Details for the file lobster_ml-1.1.418-py3-none-any.whl.
File metadata
- Download URL: lobster_ml-1.1.418-py3-none-any.whl
- Upload date:
- Size: 137.1 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? Yes
- Uploaded via: twine/6.1.0 CPython/3.13.12
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
68ca2dc1f0b081f50c5b6ad545c9e769ecfaf1026e338001ed78b1ea14c0772d
|
|
| MD5 |
1f9fb3510c7bca1621cd07c6f9009cd1
|
|
| BLAKE2b-256 |
3cdcc38820804fb533ef07fe8834e2d411df87ff966cd472c71095ab7edb4d5a
|
Provenance
The following attestation bundles were made for lobster_ml-1.1.418-py3-none-any.whl:
Publisher:
publish-packages.yml on the-omics-os/lobster
-
Statement:
-
Statement type:
https://in-toto.io/Statement/v1 -
Predicate type:
https://docs.pypi.org/attestations/publish/v1 -
Subject name:
lobster_ml-1.1.418-py3-none-any.whl -
Subject digest:
68ca2dc1f0b081f50c5b6ad545c9e769ecfaf1026e338001ed78b1ea14c0772d - Sigstore transparency entry: 1429791959
- Sigstore integration time:
-
Permalink:
the-omics-os/lobster@2201d19fcf75f9b405154412a187b58a70ed66b6 -
Branch / Tag:
refs/tags/v1.1.418 - Owner: https://github.com/the-omics-os
-
Access:
public
-
Token Issuer:
https://token.actions.githubusercontent.com -
Runner Environment:
github-hosted -
Publication workflow:
publish-packages.yml@2201d19fcf75f9b405154412a187b58a70ed66b6 -
Trigger Event:
push
-
Statement type: