Jaechang Hits@SciAgent Skills

Assemble multiple plots into ONE publication-ready multi-panel journal figure (e.g. Figure 1 with panels A, B, C). Use whenever the user asks to combine, compose, or lay out several plots as a single composite figure — newly plotted from data or from already-rendered panels the user supplies (PNG/PDF). Ask the user to pick one of two approaches: (1) redraw every panel into one unified figure using independent, tightly packed `subfigures` (each sized to its own labels, so axes need NOT align), consistent style, correctly placed panel letters, and per-panel legends/colorbars; (2) composite already-rendered PNG/PDF panels onto a mosaic canvas and add panel letters (image compositing, not plotting). Both export vector PDF + high-DPI PNG. For a SINGLE plot from a data table, use the sibling `omics-plotting` skill instead.

Interactive viewer for microscopy. Displays 2D/3D/4D arrays as Image, Labels, Points, Shapes, Tracks layers; supports annotation, plugin analysis, headless screenshots. Core visualization for Python bioimage workflows. Use ImageJ/FIJI for macro processing; napari for Python-native interactive visualization and DL segmentation review.

Python bridge to ImageJ2/Fiji for macros, plugins (Bio-Formats, TrackMate, Analyze Particles), NumPy↔ImagePlus/ImgLib2 exchange, and ImageJ Ops. Automates Fiji headlessly from Python. Use scikit-image for pure Python without Fiji plugins; napari for visualization.

omics-plotting: publication-style figure authoring for omics / bioinformatics results with matplotlib / seaborn. Read this before writing any plotting or figure code in any omics analysis — RNA-seq, proteomics, single-cell, variant, or database results — not only when a plot is explicitly requested: whenever an analysis will produce a figure, load this first and follow its recipes. Covers volcano, MA, expression / correlation heatmap, GSEA bar / dot plot, box / violin / bar / ridgeline, PCA / UMAP / t-SNE scatter, Kaplan–Meier, Manhattan / QQ / forest. Supplies a shared journal-ready style and copy-paste recipes so every figure looks like one consistent system. To combine several plots into ONE multi-panel composite figure, use the sibling `multipanel` skill.

Parse/write FCS (Flow Cytometry) files v2.0-3.1. Events as NumPy, channel metadata, multi-dataset files, CSV/FCS export. Use FlowKit for gating/compensation.

Medical image segmentation with nnU-Net's self-configuring framework — auto-selects architecture, preprocessing, training for any modality. CT, MRI, microscopy, ultrasound in 2D, 3D full-res, 3D low-res, cascade. Pipeline: convert → plan/preprocess → train (5-fold CV) → best config → predict → ensemble. Use when classical segmentation fails and annotated data exists.

Python image processing for microscopy and bioimage analysis. Read/write images, filter (Gaussian, median, LoG), segment (thresholding, watershed, active contours), measure region properties, detect features. SciPy/NumPy ecosystem. Use OpenCV for real-time video; CellPose for DL cell segmentation; napari for visualization.

Out-of-core DataFrame for billion-row data via lazy evaluation and memory-mapped files. Use when data exceeds RAM (10 GB–TB) for fast aggregation, filtering, virtual columns, and visualization without loading. Supports HDF5, Arrow, Parquet, CSV with cloud (S3, GCS, Azure). Built-in ML transformers (scaling, PCA, K-means). In-memory: polars; distributed: dask.

NEJM figure preparation: resolution (300-1200 DPI), editable vector formats (AI/EPS/SVG), in-house medical illustration policy, and strict image integrity requirements.

PNAS figure preparation: resolution (300-1000 PPI), formats (TIFF/EPS/PDF), strict RGB-only color, Arial/Helvetica fonts, italicized uppercase panel labels, automated image screening.

Computer vision for bio-image preprocessing, feature detection, real-time microscopy. Color conversion, morphology, contour/blob detection, template matching, optical flow on fluorescence/brightfield. 10-100× faster than pure Python via C++. Use scikit-image for scientific morphometry/regionprops; OpenCV for real-time, video, classical feature extraction.

Python API v2 for Opentrons OT-2/Flex liquid handlers: protocols as Python files with metadata and run(); control pipettes, labware, and modules (thermocycler, heater-shaker, magnetic, temperature). Simulate via opentrons_simulate then upload. Use PyLabRobot for vendor-agnostic scripts (Hamilton, Tecan).

eLife figure preparation: file formats (TIFF/EPS/PDF), striking image requirements (1800x900 px), figure supplement naming, and image screening policy treating selective enhancement as misconduct.

Time-to-event modeling with scikit-survival: Cox PH (elastic net), Random Survival Forests, Boosting, SVMs for censored data. C-index, Brier, time-dependent AUC; Kaplan-Meier, Nelson-Aalen, competing risks. Pipeline/GridSearchCV compatible. Use statsmodels for frequentist, pymc for Bayesian, lifelines for parametric.

Python library for single-particle tracking (SPT) in video microscopy via the Crocker-Grier algorithm. Locate particles (fluorescent spots, colloids, vesicles, cells) per frame, link into trajectories, filter short tracks, and compute MSD for diffusion analysis. 2D/3D with subpixel accuracy; reads TIF stacks, AVI, image series via pims. Use for quantitative SPT and diffusion coefficient extraction from fluorescence or brightfield video.

Fast in-memory DataFrame with lazy evaluation, parallel execution, Arrow backend. Use for tabular data in RAM (1–100 GB) when pandas is too slow. Expression API: select, filter, group_by, joins, pivots, window. Lazy mode enables predicate/projection pushdown. Reads CSV, Parquet, JSON, Excel, DBs, cloud. Larger-than-RAM: Dask; GPU: cuDF.

Nature figure preparation: resolution (300+ DPI), formats (AI/EPS/TIFF), RGB color, Helvetica/Arial fonts, lowercase panel labels, image integrity requirements.

DL cell/nucleus segmentation for fluorescence and brightfield microscopy. Pre-trained models (cyto3, nuclei, tissuenet) and a generalist flow-based algorithm segment cells without retraining. Outputs label masks for morphology and tracking. Use scikit-image watershed for rule-based; Cellpose when DL generalization across staining is needed.

WSI processing for digital pathology. Tissue detection, tile extraction (random, grid, score-based), filter pipelines for H&E/IHC. For dataset prep, tile-based DL, slide QC. Use pathml for multiplexed imaging.

Pure Python DICOM for medical imaging (CT, MRI, X-ray, ultrasound). Read/write DICOM, pixels as NumPy, edit tags, windowing (VOI LUT), PHI anonymization, build DICOM, series→3D volumes. Use histolab for WSI pathology; nibabel for NIfTI.

Science (AAAS) figure preparation: resolution (150-300+ DPI), formats (PDF/EPS/TIFF), RGB color, Myriad/Helvetica fonts, strict image manipulation policies including gamma adjustment disclosure.

Protocols and best practices for western blot quantification and analysis including band detection, normalization, and statistical methods.

Statistical visualization on matplotlib with native pandas support. Auto aggregation, CIs, grouping for distributions (histplot, kdeplot), categorical (boxplot, violinplot), relational (scatterplot, lineplot), regression (regplot, lmplot), matrix (heatmap, clustermap), grids (pairplot, FacetGrid). Use for quick statistical summaries; matplotlib for fine control; plotly for interactive HTML.

Register, segment, filter, resample 3D medical images (MRI, CT, microscopy) via SimpleITK Python; DICOM, NIfTI, multi-modal. Rigid/affine/deformable registration, threshold/region-growing segmentation, Gaussian/morph filtering, label stats, format conversion. Use to align volumes across timepoints/modalities, segment fluorescence, or convert DICOM→NIfTI.

Parallel/distributed computing for larger-than-RAM data. Components: DataFrames (parallel pandas), Arrays (parallel NumPy), Bags, Futures, Schedulers. Scales laptop to HPC cluster. For single-machine speed use polars; for out-of-core without cluster use vaex.

Python statistical modeling: regression (OLS, WLS, GLM), discrete (Logit, Poisson, NegBin), time series (ARIMA, SARIMAX, VAR), with rigorous inference, diagnostics, and hypothesis tests. Use scikit-learn for ML; statistical-analysis for test choice.

NEB-IRC activation energy pipeline for reaction barriers using GFN2-xTB and pysisyphus. Optimize reactant and product geometries, run CI-NEB path search, optimize the transition state with a Hessian, verify with IRC (one imaginary mode, endpoints matching reactant/product, single NEB maximum), and report the electronic and Gibbs barriers. Use when you need a transition state, reaction barrier, activation energy, minimum energy path, or intrinsic reaction coordinate. Covers reactant/product atom-ordering pitfalls, feasibility sizing for single-core runs, and thermochemistry corrections. Renders an IRC energy-profile plot and an animated TS imaginary-mode HTML viewer. For 2D reaction scheme drawing use rdkit-chemdraw-cdxml.

Python-based workflow manager for reproducible, scalable pipelines. Define rules with file-based dependencies; Snakemake resolves execution order and parallelism. Runs local, SLURM, LSF, AWS, GCP via profiles; per-rule conda/Singularity envs. For NGS pipelines, ML training, and multi-step file processing. Use Nextflow for Groovy dataflow or nf-core integration.

3Dmol.js WebGL molecular visualization emitted as self-contained HTML. Render structures (PDB/SDF/XYZ/MOL2/cube) with stick, sphere, cartoon, line, and surface styles; animate trajectories with a frame-delay (interval, ms) control; and animate vibrational normal modes via vibrate() from per-atom dx/dy/dz displacements or from precomputed frames. Output standalone HTML that loads 3Dmol from a CDN, with optional play/pause and speed controls. Use for transition-state imaginary-mode animations, MD or reaction-path playback, docking poses, and orbital/density isosurfaces. For static 2D chemical structure drawings use rdkit-chemdraw-cdxml; for 2D statistical plots use matplotlib or plotly.

MATLAB/GNU Octave numerical computing: matrices, linear algebra, ODEs, signal processing, optimization, statistics, scientific visualization. MATLAB-syntax examples run on both. For Python use numpy/scipy; for statistical modeling use statsmodels.

Scientific manuscript writing: IMRAD, citation styles (APA/AMA/Vancouver/IEEE), figures/tables, reporting guidelines (CONSORT/STROBE/PRISMA/ARRIVE), writing principles (clarity/conciseness/accuracy), venue-specific style. For LaTeX see companion assets.

Read, write, and edit ChemDraw CDX/CDXML files with RDKit's rdkit.Chem.rdChemDraw plus direct XML editing, always paired with a rendered PNG. Parse molecules and reactions from .cdxml/.cdx, write structures with good 2D depiction, and hand-build or modify the parts RDKit cannot write: reaction arrows, plus signs, schemes/steps, and text/labels. Use for reaction schemes, synthesis routes, mechanisms, retrosynthesis, or SI figures. Critical: RDKit writes structures only — round-tripping a reaction through a Mol silently drops arrows and text; this skill shows the XML layer that preserves them. For pure molecular analysis (descriptors, fingerprints, SMARTS) use rdkit-cheminformatics; for multi-format 3D conversion use openbabel.

smina molecular docking CLI. AutoDock Vina fork with customizable scoring functions, native SDF/MOL2/PDB ligand input, autoboxing, local energy minimization, and per-atom score breakdowns. Pipeline: receptor PDBQT prep -> ligand prep (RDKit/OpenBabel) -> dock via autobox or explicit grid -> rescore/minimize with custom scoring -> rank poses by affinity. Choose smina over Vina when you need custom scoring terms (--custom_scoring), local optimization of an existing pose (--local_only), per-atom contributions (--atom_term_data), or SDF/MOL2 ligands without manual PDBQT conversion. For unknown binding sites use diffdock; for the Python-bindings/Vinardo workflow use autodock-vina-docking.

mdtraj molecular dynamics trajectory analysis (Python). Reads DCD/XTC/TRR/NetCDF/H5/PDB topologies and trajectories; computes RMSD vs time, radius of gyration, per-residue RMSF, residue-residue contact frequency maps, phi/psi torsions for Ramachandran plots (general + Gly/Pro), and 8-state DSSP secondary structure. Modules: trajectory I/O, geometry (distances/angles/dihedrals), structural analysis (RMSD/Rg/RMSF/SASA), contacts, hydrogen bonds, secondary structure (DSSP), NMR observables. For broader atom-selection grammar use mdanalysis-trajectory; for running MD simulations use OpenMM/GROMACS.

AiZynthFinder retrosynthetic route planning (CASP) from AstraZeneca Molecular AI. Monte Carlo tree search guided by a template-based neural expansion policy recursively disconnects a target SMILES until precursors are found in a purchasable stock. Covers config.yml (v4 format), aizynthcli batch screening, the AiZynthFinder/AiZynthExpander Python API, one-step disconnections, custom stocks via smiles2stock, scorers, Retro*/breadth-first/DFPN search alternatives, and reading output.json.gz / trees.json. Use for synthesis route planning, synthesizability screening, and building-block/precursor search. For reaction barriers use neb-irc-activation-energy; for 2D reaction scheme drawing use rdkit-chemdraw-cdxml.