dailycafi

Look up any metabolite in the Human Metabolome Database (HMDB) via REST API across 220,000+ entries. Use when: user asks 'what is this metabolite', needs an HMDB ID, wants metabolite pathways or disease associations, queries a metabolite database, or needs cross-references to KEGG/PubChem/ChEBI. Triggers: metabolite lookup, HMDB search, metabolite properties, metabolite spectra, metabolite biomarker, compound information, metabolite concentration, biofluid metabolites, serum metabolites, urine metabolites.

Search and download public metabolomics study data from EMBL-EBI MetaboLights via REST API (2,800+ studies). Use when: user wants to find metabolomics studies, download study data, retrieve ISA-Tab files, access public metabolomics datasets, or get curated metabolite annotations. Triggers: MetaboLights, MTBLS, download study data, public metabolomics data, ISA-Tab, find metabolomics studies, metabolite assignment file, MAF file, deposited metabolomics, open-access metabolomics repository.

Query the NIH Metabolomics Workbench REST API across 4,200+ studies for metabolite data, RefMet nomenclature, and spectral searches. Use when: user asks about Metabolomics Workbench, needs RefMet standardized names, performs m/z search against a database, wants NIH metabolomics data, or retrieves gene-metabolite associations. Triggers: metabolomics workbench, NIH metabolomics, RefMet, m/z search, exact mass search, metabolite structure search, PubChem CID lookup, study metadata, GC-MS/LC-MS/NMR public data.

Constraint-based metabolic modeling with COBRApy. Flux balance analysis (FBA), flux variability (FVA), gene knockouts, flux sampling, production envelopes, and gap filling on genome-scale SBML models. Use when: predicting growth rates, optimizing metabolic fluxes, screening gene deletions, or building metabolic models. Triggers: FBA, FVA, COBRA, metabolic model, SBML, flux analysis, gene knockout simulation, metabolic engineering, growth prediction.

Interpret clinical metabolomics results for inborn errors of metabolism (IEM) screening, newborn screening, and diagnostic reporting. Use when: user has clinical metabolite panels, needs IEM differential diagnosis, wants to analyze acylcarnitine profiles or amino acid panels, or calculate z-scores against reference ranges. Triggers: newborn screening, IEM, inborn error of metabolism, acylcarnitine, amino acid panel, organic acid analysis, clinical diagnosis, tandem MS screening, PKU, MCADD, maple syrup urine disease, clinical metabolomics, reference range, z-score.

Analyze lipidomics data for lipid species identification, quantification, and pathway interpretation using LipidSearch, MS-DIAL, and LIPID MAPS. Use when: user has lipidomics data, needs lipid class annotation, wants to analyze sphingolipids/phospholipids/fatty acids, or interpret lipid-specific pathways. Triggers: lipid species, lipidomics, sphingolipid, phospholipid, fatty acid, ceramide, triglyceride, lipid class, LIPID MAPS, LipidSearch, lipidr, lipid annotation, chain composition, lipid profiling, phosphatidylcholine, PE, PC, SM.

Annotate and identify metabolomics features by matching m/z, retention time, and MS/MS spectra against databases. Use when: user has a feature table and wants compound IDs, needs to annotate m/z values, assign metabolite identities with confidence levels, or match features against HMDB/METLIN/MassBank. Triggers: identify features, m/z annotation, compound identification, metabolite ID, putative annotation, MSI confidence levels, adduct matching, ppm tolerance, neutral mass search, what compound is this m/z.

Process metabolomics data with MS-DIAL for peak detection, alignment, annotation, and export to feature tables. Use when: user has MS-DIAL output files, wants to import MS-DIAL results into R/Python, needs peak alignment from MS-DIAL, or prefers GUI-based LC-MS preprocessing. Triggers: MS-DIAL, MS-DIAL output, MSDIAL, peak alignment, MS-DIAL export, GUI preprocessing, alternative to XCMS, MS-DIAL console mode, metabolomics feature table from MS-DIAL.

PyTorch training recipes for all neural network domains. Use when training or fine-tuning models, debugging loss spikes or NaN gradients, resolving OOM errors, choosing model architectures, selecting optimizers, tuning learning rates, or optimizing GPU throughput. Covers LLMs, vision, diffusion, medical imaging, protein/drug discovery, spatial omics, and genomics. Do NOT use for pandas data analysis, API serving, unit testing, or scRNA-seq differential expression.

Write publication-ready biomedical research papers with IMRAD structure, CNS-style figures, LaTeX generation, and PubMed-verified citations. Use when user mentions writing a paper, drafting a manuscript, IMRAD, journal names (Nature, Science, Cell, NEJM, Lancet, JAMA), reporting guidelines (CONSORT, STROBE, PRISMA, ARRIVE), reviewer comments, cover letters, or converting lab results into manuscript form. Also triggers for generating publication figures, LaTeX compilation, multi-omics paper checklists, and citation verification.

Plan, run, resume, audit, and report proteomics database searches on Linux with FragPipe/MSFragger and compatible validation or quantification tools. Use for DDA or DIA, label-free or isobaric workflows, total proteome or PTM-enriched data, protein/peptide/PTM identification, targeted protein or residue checks, SSH/tmux execution, FASTA preparation, FDR review, quantification, and publication-ready tables, figures, and source-data packages. Includes PTM-specific branches without assuming any particular protein, gene, modification, organism, or sample design.

Plan, run, review, and report defensible biomolecular docking and molecular dynamics workflows. Use for protein-ligand docking, docking-pose review, docking-to-MD validation, protein-only MD, membrane-protein MD, multimer modeling, mutation/redocking/MD comparisons, ligand or lipid parameterization decisions, trajectory analysis, free-energy escalation, and cases where MD should be avoided or caveated because the biological assembly, membrane state, protonation, ligand parameters, or starting pose is not reliable.

Develop teaching cases for business schools in collaboration with case authors. Covers the full lifecycle from topic evaluation through writing, classroom testing, to journal submission. Includes a 6-stage iterative workflow, 5-dimension topic evaluation matrix, rubric-aligned quality checks, and teaching note design. Use when user mentions teaching case, case development, case writing, teaching note, topic evaluation for business cases, interview preparation for case research, case submission, or case rubric review. Also triggers when user shares interview transcripts, case drafts, reviewer feedback, or asks to evaluate a case topic. Supports Chinese and English cases.

处理用户反馈文档（docx）并修改项目代码。当用户提到"用户反馈"、"修改需求"、"反馈文档"、拖入 docx 文件、或指向 feedback-docs 目录时触发。