Drug & Clinical Pipeline Explorer

Welcome to the Ambrosigen Drug & Clinical pipeline Explorer (ambrosigen.com/tools-and-resources/drug-and-clinical-trial-explorer), a browser-based clinical pipeline exploration tool. It lets researchers, clinicians, investors and drug developers explore the clinical development pipeline, track trials, see how far each drug has progressed and whether it is still moving, and browse target/mechanism data.

The Explorer runs inside your web browser. It requires no accounts or logins, and your filters and searches are never sent to a server.

Before you start, two things worth knowing: currently, the dataset covers small molecules only (biologics might be added later), and trial data comes from ClinicalTrials.gov, so development registered only outside it is not visible. Section 6 covers the limitations in full.

Search approved and investigational drugs, their clinical trials and development status in a full-window explorer.

Open the explorer

1. Core Capabilities: What Can It Do for You?

The Drug Explorer is designed to answer complex clinical development and competitive intelligence questions instantly:

  • Pipeline Landscaping: Explore clinical pipelines by target gene, mechanism of action, therapeutic area, specific indication, or drug class (WHO ATC codes).
  • Progress Tracking: See which drugs are advancing, holding at their deepest phase, have fallen back below it, or have no active development — independently of whether they are approved. Where a programme stopped, the Trials tab shows at which phase and the recorded reason.
  • Timeline and Development Metrics: Analyze clinical development times, phase transition history, and trial start trends over time.
  • Regulatory Status Audits: Cross-reference regulatory achievements such as FDA New Molecular Entity (NME) approvals, EMA marketing authorizations, PMDA listings, and special designations (e.g., Orphan Drug, Breakthrough Therapy, Fast Track).
  • Trial Linkage and Arm Role Audits: Distinguish trials where a drug is the primary subject under evaluation (investigational) from trials where it is used as a reference treatment (comparator) or standard-of-care baseline (background).
  • Evidence Summaries: View a detailed record of all trials linked to a drug, along with published results, primary/secondary outcomes, study enrollment, and eligibility criteria (including age and gender limits).

2. Platform Architecture: The 6 Data Sources

The Drug Explorer merges and harmonizes six authoritative public and curated data sources to create a unified, single source of truth:

SourceWhat it ContributesWhy It Matters
ChEMBLPrimary compound spine, standard chemical structures (SMILES/InChIKey), physical/chemical properties, first approval years, and official safety/withdrawal warnings.Established the primary database of clinical-stage small molecules and physical structures.
AACT (Duke CTTI / ClinicalTrials.gov)Trial-level metadata (phases, status, start/completion dates, enrollment, eligibility) and experimental arm layouts.Drives the primary trial-linkage engine, resolving precisely which compounds were tested.
Open TargetsTargets, mechanisms of action, disease indications, and a curated trial stop-reason classification overlay.Provides the scientific vs. administrative classification for stopped trials, which does not exist in raw CT.gov.
FDA NME CompilationDay-precision approval dates, original applicants, and special regulatory designation flags.Authoritative for FDA approval status and regulatory benefits.
EMA Medicines ReportEU marketing authorizations, dates, applicants, and European special designations.Backfills EU-approved drugs that are structurally absent from US-centric datasets.
KEGG br08328Multi-agency approval backfills (PMDA/FDA/EMA) and additional anatomical therapeutic chemical (ATC) mappings.Extends non-US approval coverage and drug classification.

3. How to Use the Interface

The application is structured into three primary tabs that update dynamically as you apply filters:

A. The Filter Bar (Global Controls)

The filters at the top of the application apply across all tabs:

  • Search Input: Search by compound name, research codes (e.g., MK-1234), brand names, ChEMBL ID, or ClinicalTrials.gov NCT number.
  • Target/Mechanism Filters: Filter by targeted gene symbol (e.g., EGFR), biochemical action (e.g., Inhibitor, Antagonist), or cellular pathway category.
  • Sponsor/Developer Filters: Type-ahead filter on the normalised sponsor name (legal-entity spelling variants folded together).
  • Regulatory Designations: Tick any combination of First-in-class, Orphan Drug, Breakthrough Therapy, Fast Track, Accelerated Approval, QIDP, Withdrawn and Black-box warning. Ticking several requires all of them at once — Orphan and Breakthrough narrows the set rather than widening it. Every designation a drug holds is listed in its detail panel regardless of what is ticked here.
  • Drug Class (USAN stem): Type-ahead over the international drug-naming stems — -tinib (tyrosine kinase inhibitors), -ciclib (CDK inhibitors), -vir (antivirals), and 664 others. Present on 7,150 of 12,813 drugs, making it the highest-coverage class signal in the dataset. Matching is case-insensitive over both the stem and its definition, so kinase works as well as -tinib. Drugs with no stem assigned are excluded while the box has text in it.
  • Route: Oral, parenteral or topical. A drug can have several and matches under any of them. In pooled mode every salt form’s routes count, since route belongs to the formulation rather than the molecule.
  • Chemistry: Prodrug (inactive as administered, converted to the active form in the body) and Natural product (structure derived from a natural source rather than being fully synthetic). Ticking both requires both.
  • Status Toggles:
    • Drug Status: Two independent axes. Approval (approved / not approved, with an approval-date range and a “withdrawn only” option) and Progress (Advancing, Active at peak, Fell back, No active development, No trial data). Because they are independent you can ask for combinations a single status could not express — for example approved drugs that have fallen back. Also here: a last-activity range, and “has an ongoing trial at Phase N”.
    • Trial Status: Filter by individual trial status (Active, Completed, Stopped, Expanded Access, or Unknown).
  • Data Aggregation Toggle (Pooled vs. Per-Formulation):
    • Pooled Mode (Default): Aggregates salts, esters, and prodrugs into a single parent active moiety (e.g., sildenafil citrate and sildenafil are pooled under “Sildenafil”). This prevents duplicate counting and shows the true clinical evidence weight.
    • Per-Formulation Mode: Views every registered salt form, solvate, and derivative as a distinct individual record.

B. The “Drugs” Tab (Compound Inventory)

This tab shows a table listing all clinical-stage compounds matching your active filters.

  • Stage (Highest Phase): The highest phase of clinical development the drug has ever entered. Phase 4 indicates an approved status.
  • Trials Count: Consolidates all linked trials. A custom hover-explained role bar breaks down these trials into their functional roles:
    • Teal: Investigational trials on an interventional study (this counts toward the drug’s core trial-evidence total).
    • Purple: Reference comparator trials.
    • Grey: Background / backbone regimen trials.
    • Slate: Observational / Expanded Access trials (not counted toward core totals).
  • Last Activity: The date of the most recent CT.gov trial update for this drug, signaling whether development is active or dormant.
  • Detail Panel (Clicking a Drug): Clicking on any drug row opens a side-drawer or detailed profile including:
    • Chemical structure rendering (SMILES/InChIKey) and physical properties (MW, LogP, Polar Surface Area).
    • Comprehensive list of regulatory approvals and designations across FDA, EMA, and KEGG.
    • Curated target, action mechanism, and Reactome pathway lists.
    • Historical timeline of clinical trial involvement.
    • Official safety/withdrawal warnings.

C. The “Trials” Tab (Study Details)

This tab lists every individual drug-trial link matching your criteria. It is powered by the local DuckDB database, enabling instant search and filtering.

  • Trial ID (NCT): Clicking this opens the official clinicaltrials.gov record page.
  • Role & Regimen: Shows whether the drug was the investigational subject, comparator, or background backbone in this study, and whether it was administered as a monotherapy or part of a combination regimen.
  • Status & Bucket: Displays the current CT.gov trial status alongside its normalized lifecycle bucket (Ongoing, Completed, Stopped, Stalled, Expanded Access, Unknown).
  • Why Stopped: Surfacings the curated reason why a trial was terminated early (Efficacy/Safety, Administrative, Regulatory, or Unclear) alongside the raw text description filed by the sponsor.
  • Results & Outcomes: Lists documented primary and secondary endpoints with their actual measured values (where available).
  • Eligibility Criteria: Summarizes age ranges, gender requirements, and healthy volunteer rules.

D. The “Charts” Tab (Visual Analytics)

Generates high-resolution visualizations based on your filtered active cohort:

  • Clinical Funnel: Attrition as drugs transition from Phase 1 to approval, scoped to drugs with a resolved outcome so unfinished programmes do not inflate it.
  • Success Rates: Historical transition rates within a chosen entry-date cohort. Read the caveat in section 6 before quoting these numbers.
  • Approval × Progress: The two independent axes crossed — including “Approved · Fell back”, an approved drug whose development has retreated, which no single-status model can express.
  • Development Timelines: Duration from first-in-human trial to regulatory approval.
  • Why Trials Stop: Scientific failures (safety/efficacy) vs. administrative halts (logistics, recruitment, business decisions). This is a trial-level view: one stopped trial does not mean the drug was discontinued.

4. Key Data Models & Definitions

To get the most out of the Drug Explorer, it helps to understand its key data definitions and logic:

Approval and Progress — two independent facts

A drug is described by two separate things at once, not by one combined status. This matters: a drug can be approved and have retreated from the phase it once reached, and older versions of this tool could not say that.

Approval is a regulatory fact. Established from FDA, EMA, or ChEMBL/KEGG br08328 records. KEGG and ChEMBL are data sources, not regulators — where the approving authority is known it is shown explicitly; ChEMBL-sourced approvals do not identify a country. PMDA (Japan) is currently dormant and contributes no approvals, and there is no Chinese (NMPA) source, so some drugs approved in China correctly show as not approved here.

Progress is momentum through the development pipeline. It compares the deepest phase with a currently ongoing trial against the deepest phase the drug has already resolved (completed, stopped or stalled):

ProgressMeaning
AdvancingRunning a trial deeper than anything it has resolved — entering new ground.
Active at peakRunning trials at the deepest phase it has ever reached.
Fell backStill running trials, but only below its peak — e.g. a Phase 3 programme that stopped and is now back in Phase 1.
No active developmentNo trial currently running at a development phase (Phase 0/1–3). Includes drugs whose only ongoing trials are Phase 4 post-marketing studies.
No trial dataNo trial studies this drug as the investigational agent at all — a statement about our coverage, not an outcome.

Progress is capped at Phase 3, because Phase 4 is post-marketing activity rather than development. It never reads a stop reason, and a stopped trial can never outrank live work.

“No active development” does not mean “failed.” A programme that completed and is awaiting readout looks identical to one that was abandoned. Where a drug stopped, the Trials tab shows which phase and why.

Trial Role Classifications

To prevent reference compounds (e.g., standard chemotherapy backbones or comparative controls) from artificially inflating a drug’s developmental profile, the pipeline runs an automated arm-role classifier on AACT’s design layouts:

  • Investigational: The drug is the core agent being tested; its presence varies across arms.
  • Comparator: The drug is used as an active control/reference arm to test against another agent.
  • Background: The drug is given to all arms as a standard basal therapy (e.g., metformin in diabetes trials).
  • Inferred: If arm-level data is missing, the system infers an investigational role only if (a) it is the only drug intervention listed, or (b) its name appears as a full token in the trial’s official title.

Curated Stop Reasons

When an interventional trial is stopped early, its why_stopped text is run through a high-severity taxonomic classifier:

  • Efficacy/Safety (Scientific): Trial terminated due to lack of efficacy, safety concerns, or side effects. This represents a clinical hurdle.
  • Administrative: Terminated due to recruitment issues, logistical constraints, business decisions, funding cuts, or COVID-19.
  • Regulatory: Halted due to an explicit regulatory agency block or safety hold.
  • Unclear: Stated text is too vague to categorize or missing details.

5. Scope and known limitations

Read this before drawing conclusions from the numbers.

  • Small molecules only. Biologics — antibodies, recombinant proteins, peptides, vaccines, cell and gene therapies — are not in the dataset. Searching for pembrolizumab, adalimumab or semaglutide returns nothing, by design rather than by omission.
  • Trials are US-registered. ClinicalTrials.gov is the only trial registry used, so development conducted entirely outside it is invisible. This is the main reason a drug can show “No trial data” while being genuinely active elsewhere.
  • Partial curation coverage. Target ~28%, action type ~33%, pathways ~28%, approved indications ~21%. A blank cell is a curation gap, not evidence that the drug has no target or mechanism.
  • Success rates read far higher than the published literature. The Charts tab computes roughly 50% Phase-1-to-approval against a widely published figure nearer 10%. The gap is registry coverage, not a claim that these drugs succeed more often: where coverage is good the method agrees with reality (our Phase 3 → approval is close to the published figure), and it is weakest exactly where real attrition is worst. Use these charts to compare cohorts within the tool, not as absolute industry rates.
  • Cross-therapeutic-area comparison is not supported. Most drugs are studied in several indications, and a drug that failed in five and advanced in one still counts as advancing. This can invert area-level comparisons, so they are deliberately not offered as a headline.

Enjoy exploring the clinical pipeline. For bug reports, feature requests or dataset issues, please get in touch via ambrosigen.com.

4. Key Data Models & Definitions

To get the most out of the Drug Explorer, it helps to understand its key data definitions and logic:

Approval and Progress — two independent facts

A drug is described by two separate things at once, not by one combined status. This matters: a drug can be approved and have retreated from the phase it once reached, and older versions of this tool could not say that.

Approval is a regulatory fact. Established from FDA, EMA, or ChEMBL/KEGG br08328 records. KEGG and ChEMBL are data sources, not regulators — where the approving authority is known it is shown explicitly; ChEMBL-sourced approvals do not identify a country. PMDA (Japan) is currently dormant and contributes no approvals, and there is no Chinese (NMPA) source, so some drugs approved in China correctly show as not approved here.

Progress is momentum through the development pipeline. It compares the deepest phase with a currently ongoing trial against the deepest phase the drug has already resolved (completed, stopped or stalled):

ProgressMeaning
AdvancingRunning a trial deeper than anything it has resolved — entering new ground.
Active at peakRunning trials at the deepest phase it has ever reached.
Fell backStill running trials, but only below its peak — e.g. a Phase 3 programme that stopped and is now back in Phase 1.
No active developmentNo trial currently running at a development phase (Phase 0/1–3). Includes drugs whose only ongoing trials are Phase 4 post-marketing studies.
No trial dataNo trial studies this drug as the investigational agent at all — a statement about our coverage, not an outcome.

Progress is capped at Phase 3, because Phase 4 is post-marketing activity rather than development. It never reads a stop reason, and a stopped trial can never outrank live work.

“No active development” does not mean “failed.” A programme that completed and is awaiting readout looks identical to one that was abandoned. Where a drug stopped, the Trials tab shows which phase and why.

Trial Role Classifications

To prevent reference compounds (e.g., standard chemotherapy backbones or comparative controls) from artificially inflating a drug’s developmental profile, the pipeline runs an automated arm-role classifier on AACT’s design layouts:

  • Investigational: The drug is the core agent being tested; its presence varies across arms.
  • Comparator: The drug is used as an active control/reference arm to test against another agent.
  • Background: The drug is given to all arms as a standard basal therapy (e.g., metformin in diabetes trials).
  • Inferred: If arm-level data is missing, the system infers an investigational role only if (a) it is the only drug intervention listed, or (b) its name appears as a full token in the trial’s official title.

Curated Stop Reasons

When an interventional trial is stopped early, its why_stopped text is run through a high-severity taxonomic classifier:

  • Efficacy/Safety (Scientific): Trial terminated due to lack of efficacy, safety concerns, or side effects. This represents a clinical hurdle.
  • Administrative: Terminated due to recruitment issues, logistical constraints, business decisions, funding cuts, or COVID-19.
  • Regulatory: Halted due to an explicit regulatory agency block or safety hold.
  • Unclear: Stated text is too vague to categorize or missing details.

5. Scope and known limitations

Read this before drawing conclusions from the numbers.

  • Small molecules only. Biologics — antibodies, recombinant proteins, peptides, vaccines, cell and gene therapies — are not in the dataset. Searching for pembrolizumab, adalimumab or semaglutide returns nothing, by design rather than by omission.
  • Trials are US-registered. ClinicalTrials.gov is the only trial registry used, so development conducted entirely outside it is invisible. This is the main reason a drug can show “No trial data” while being genuinely active elsewhere.
  • Partial curation coverage. Target ~28%, action type ~33%, pathways ~28%, approved indications ~21%. A blank cell is a curation gap, not evidence that the drug has no target or mechanism.
  • Success rates read far higher than the published literature. The Charts tab computes roughly 50% Phase-1-to-approval against a widely published figure nearer 10%. The gap is registry coverage, not a claim that these drugs succeed more often: where coverage is good the method agrees with reality (our Phase 3 → approval is close to the published figure), and it is weakest exactly where real attrition is worst. Use these charts to compare cohorts within the tool, not as absolute industry rates.
  • Cross-therapeutic-area comparison is not supported. Most drugs are studied in several indications, and a drug that failed in five and advanced in one still counts as advancing. This can invert area-level comparisons, so they are deliberately not offered as a headline.

Enjoy exploring the clinical pipeline. For bug reports, feature requests or dataset issues, please get in touch via ambrosigen.com.

Are you looking for someone with expertise in medicinal or synthetic chemistry?

Do you need a custom synthesis of small molecules?