3. Data, Science & AI

Provides tools for building and training Graph Neural Networks (GNNs) using PyTorch Geometric, including node classification, link prediction, and molecular property prediction.

Queries NCBI Gene database via E-utilities to retrieve gene information including RefSeqs, GO terms, locations, and phenotypes for annotation and functional analysis.

Provides general rules and best practices for generating effective prompts in large language model (LLM) interactions.

Guides structured experiment configuration in machine learning using Hydra and YAML to ensure reproducibility and clarity.

Ensures AI verifies information before presentation to prevent speculation and maintain factual accuracy.

Accesses RCSB PDB database to search, download, and retrieve metadata for 3D protein/nucleic acid structures used in structural biology and drug discovery research.

Interactive visualization library for creating web-based charts with hover, zoom, and pan features, ideal for dashboards and exploratory data analysis.

High-performance in-memory DataFrame library for fast data processing, ETL pipelines, and pandas replacement with lazy evaluation and Apache Arrow backend.

Enables NLP, computer vision, and multimodal tasks using pre-trained transformer models for generation, classification, and fine-tuning.

Assists with database design, implementation, and optimization using SQL, NoSQL, Prisma, and Supabase.

Provides pandas data manipulation rules: method chaining, loc/iloc, and groupby operations for efficient data processing.

Python wrapper for RDKit simplifying drug discovery tasks including SMILES parsing, standardization, and molecular descriptors with native RDKit object support.

Applies medicinal chemistry filters including Lipinski, Veber, PAINS, and structural alerts to prioritize compounds and filter chemical libraries.

Queries PubChem database for chemical compounds via PUG-REST API or PubChemPy, supporting name/CID/SMILES searches and retrieving properties, similarity, and bioactivity data.

Defines operational modes (planning, editing, interactive) for AI agents to self-regulate tool usage based on current development phase.

Retrieves enzyme kinetic parameters, reaction equations, and organism data from BRENDA database via SOAP API for biochemical research and metabolic pathway analysis.

High-performance toolkit for genomic interval analysis, supporting BED files, coverage tracks, and ML tokenization in computational genomics.

Provides Python-based access to microscopy data for image retrieval, pixel analysis, ROI management, and batch processing in high-content screening workflows.

Python toolkit for graph analysis, visualization, and algorithmic computation of network structures and relationships.

Cloud-based quantum chemistry platform with Python API for molecular property prediction, protein-ligand docking, and AI-driven computational workflows.

Infers gene regulatory networks from transcriptomics data (bulk and single-cell RNA-seq) using scalable algorithms like GRNBoost2 and GENIE3.

Performs differential gene expression analysis on RNA-seq data using DESeq2 methodology, including Wald tests, FDR correction, and visualization plots.

Enables genomics pipeline development, data management, and workflow execution on DNAnexus using dxpy SDK for FASTQ/BAM/VCF processing.

Toolkit for ESM3 and ESM C protein language models enabling sequence, structure, function prediction, and protein design via local or cloud-based inference.