Drug Discovery Informatics

Create a comprehensive target validation report for [protein target]:

Target: [protein name and gene symbol]
Disease area: [indication(s) of interest]
Modality under consideration: [small molecule inhibitor, degrader, antibody, etc.]

Assess across validation pillars:
1. Genetic evidence — GWAS hits, Mendelian genetics, loss-of-function studies
2. Expression data — disease vs. normal tissue expression, cell-type specificity
3. Functional biology — pathway role, known substrates/interactors, knockout phenotypes
4. Published preclinical evidence — animal models, in vitro disease models
5. Clinical precedent — has this target been drugged before? What were the outcomes?
6. Druggability assessment — structural data, known binding sites, tool compounds
7. Safety considerations — essential function in normal physiology, predicted on-target toxicity
8. Competitive landscape — who else is pursuing this target and at what stage

Rate each pillar (strong / moderate / weak / no data) and provide an overall target validation score with rationale. Identify the single biggest risk and the most compelling piece of evidence.

Analyze bioactivity data for [target protein] from ChEMBL or internal assay data:

Target: [protein name, ChEMBL target ID if known]
Data scope: [all public data, specific assay types, date range]
Compound set: [all published, specific chemotypes, internal series]

Analyze:
1. Data overview — total measurements, assay types (IC50, Ki, Kd, EC50), year distribution
2. Potency landscape — distribution of values, most potent compounds, potency cliffs
3. Assay reproducibility — consistency across labs and assay formats
4. Structure-activity relationships (SAR) — what structural features drive potency
5. Selectivity data — if available, how selective are the most potent hits against related targets
6. Chemical matter quality — drug-likeness (MW, logP, TPSA), synthetic accessibility
7. Publication context — which papers report the key compounds and what were the study goals

Provide:
- Summary statistics table (median, mean, range by assay type)
- Top 10 most potent compounds with structures (SMILES) and key properties
- SAR hypotheses with supporting evidence
- Recommendations for hit-to-lead optimization priorities

Optimize the Design-Make-Test-Analyze (DMTA) cycle for [drug discovery program]:

Program context:
- Target: [protein target]
- Modality: [small molecule, PROTAC, molecular glue, etc.]
- Current cycle time: [days per DMTA iteration]
- Bottleneck: [design, synthesis, assay, analysis]
- Team size: [number of medicinal chemists, biologists, informaticians]

For each DMTA stage, assess and recommend:

Design:
- Computational tools in use vs. available (docking, FEP, ML models, generative chemistry)
- How design hypotheses are prioritized and documented
- Integration of ADMET predictions early in design

Make:
- Synthesis planning and route optimization
- Parallel synthesis vs. focused medicinal chemistry
- Compound registration and tracking workflow

Test:
- Assay cascade structure (primary, secondary, selectivity, DMPK)
- Data flow from instruments to database
- Turnaround time per assay tier

Analyze:
- How results are curated, validated, and flagged
- SAR visualization and multiparameter optimization tools
- Decision-making framework for advancing vs. deprioritizing series

Provide specific recommendations to reduce cycle time by [target percentage], with tool and process changes for each stage.

Identify opportunities for a new drug to improve upon existing therapies in [indication]:

Disease area: [condition]
Current standard of care: [approved drugs/treatments]
Proposed mechanism: [new drug's mechanism of action]
Proposed modality: [e.g., oral small molecule vs. injectable biologic]

Analyze dimensions of potential improvement:
1. Efficacy — where does the current SOC fall short? Response rates, durability, specific endpoints
2. Safety — what are the dose-limiting toxicities or long-term safety concerns of current therapies?
3. Convenience — route of administration, dosing frequency, monitoring requirements
4. Patient access — cost, insurance coverage, geographic availability
5. Mechanism breadth — does the new approach address pathways the current SOC misses?
6. Resistance — are patients developing resistance to current therapies? Can the new approach overcome it?
7. Combination potential — can the new drug combine with SOC for additive/synergistic benefit?
8. Underserved populations — patient subgroups not well served by current options

For each dimension, rate the improvement opportunity (high / moderate / low) and provide evidence. Summarize the overall value proposition and target product profile.

Assess the selectivity profile of [compound or compound series]:

Compound: [name or ID]
Primary target: [intended target]
Related targets to assess: [list of off-target proteins to evaluate]
Available data: [selectivity panel results, published data, predicted]

Analyze:
1. On-target potency — IC50/Ki/Kd at the primary target
2. Off-target activity — potency at each related target
3. Selectivity ratios — fold selectivity for primary vs. each off-target
4. Structural basis — if structural data exists, what drives selectivity or lack thereof
5. Functional consequences — what would off-target activity mean pharmacologically?
6. Safety implications — which off-targets are associated with known toxicities
7. Species differences — does selectivity profile differ between human and preclinical species

Provide:
- Selectivity matrix table (compound × target × potency)
- Traffic light assessment (green/amber/red) for each off-target
- Recommended selectivity assays to fill data gaps
- Design hypotheses to improve selectivity if needed

Design a systematic literature review strategy for [research question]:

Research question: [specific question to answer]
Therapeutic area: [disease/target/modality]
Scope: [systematic review, scoping review, rapid evidence assessment]
Timeline: [available time for the review]

Define:
1. Search strategy — databases (PubMed, Embase, Cochrane, bioRxiv), search terms, Boolean logic, date range
2. Inclusion/exclusion criteria — study types, populations, outcomes, languages
3. Screening process — title/abstract screening, full-text review, number of reviewers
4. Data extraction template — what fields to capture from each study
5. Quality assessment — which risk-of-bias tool to use and how to apply it
6. Synthesis approach — narrative, meta-analysis, vote counting, or qualitative
7. PRISMA flow diagram template

Provide the complete search strings for PubMed ready to execute, and a data extraction spreadsheet template with column definitions.