What changed in GenRx since the September manual.
Use these operating instructions with your organization's current permissions and release. This describes product workflows, not a claim of completed customer GxP validation, regulator acceptance or universal feature availability.
1. Program OS: select, create and safely edit your programme
- Sign in and open My Programs.
Open Program OS and explicitly select an authorized programme. Do not assume the first listed programme was selected automatically. If none is available, use the separate New programme action if you have permission.
- Verify identity before entering scientific work.
Check the product, strength, dosage form, market, canonical revision and selected programme. Product identity and controlled scientific revisions are protected; ordinary planning-metadata edits must not silently change them.
- Edit planning metadata using its dedicated Save action.
New programme and Save planning metadata are different actions. Save against the current version; the application must confirm the persisted change through a fresh authoritative read-back. If a conflict, error or missing receipt appears, refresh and resolve it rather than assuming success.
- Interpret readiness from accepted evidence.
Drafts, calculations, unsaved edits and unreviewed module outputs do not advance approved programme progress. Check blockers, the current gate, reviewer authority and the next controlled action.
- Recover deliberately.
An empty scoped list, anonymous sign-in prompt, expired session, permission denial and service failure are distinct states. Sign in, change permitted scope or retry as indicated; never create substitute data to force a populated view.
2. Scientific Requirements: US FDA, EU/EMA and India/CDSCO
Open the Scientific Requirements Workbench for a programme-first, market-specific view of ten connected scientific workstreams. The workbench is designed around source-linked requirements and server-controlled evidence states, not three independent science engines.
- Choose a verified programme and canonical revision.
The selector provides actual accessible programmes. With no verified programme, Explore mode is for navigation only: do not infer product readiness or save governed evidence against an illustrative context.
- Select India, EU or US.
Review the applicable workstream and its Evidence, Sources and Requirement views. Confirm the product's actual strength, comparator, dosage form, current official source and jurisdictional applicability.
- Compare markets carefully.
Use Compare markets to inspect the server-authoritative evidence matrix. Different statuses do not prove scientific equivalence, transferable studies or regulatory permission to reuse evidence.
- Attach evidence through the authorized programme.
Evidence handoff requires a verified programme, current product revision and correct market. Saved evidence starts as draft and unreviewed; typed identifiers and source pointers are not validation or reviewer approval.
- Respect initial scientific scope.
Jurisdiction guidance is initially oriented to immediate-release tablets and capsules. Do not apply these prompts as validated rules for modified-release, injectable, inhaled, topical, biologic or other complex products.
Review current FDA, EMA and CDSCO sources, product-specific guidance, applicable pharmacopeial monographs and the evidence with qualified scientific and regulatory staff before drawing conclusions.
3. Evidence Review Center and controlled decision packets
Preserve canonical module identity, method/version, source, product revision, assumptions, evidence class and limitations. Handoffs are draft and unreviewed by default.
Review immutable versions, field differences and current versus stale status before taking action.
Approve, reject or request changes using the authenticated, tenant- and programme-scoped reviewer workflow; a displayed reviewer name is not an electronic signature.
A fetched or imported source is not automatically current, applicable or accepted. A changed live source can invalidate the accepted snapshot.
Use only current approved evidence and an accepted, matching source snapshot. Stale, rejected, superseded and changes-requested evidence must not silently enter the packet.
The P2 Development Workbench also exposes decision support for vBE/PBBM credibility, QbD/DoE, mechanistic scale-up, stability statistics, ANDA/eCTD technical preflight, Source Watch impact, complex-generics archetypes, portfolio economics and patent/exclusivity planning. These analyses inform governed decisions; they are not a validated eCTD publisher, shelf-life authorization, bioequivalence ruling or freedom-to-operate legal opinion.
4. BayesPeaQ Quality: current role and SOP configuration
My Work is the starting point for assigned Quality work. Team and oversight views depend on active Quality roles; Tenant Admin alone does not grant QMS authority. The Role Configuration Studio provides controlled navigation and responsibility configuration through Role Workspaces, Workflow Responsibilities, Approval Inbox and Version History.
How do organization roles relate to Role Studio?
Use the same tenant-bound, versioned role catalogue for organization assignments and Studio configuration. Stage-owner selectors deliberately show only eligible roles as selectable. Non-eligible roles may appear disabled with a reason; that is a governance rule, not necessarily a missing-role defect. The approved role catalogue and tenant assignments remain separate from platform-administrator authority.
How do I add or edit an SOP activity?
In the authorized Responsibilities or Quality Configuration view, select the approved workflow and lifecycle stage, then define the activity, responsible eligible Quality role, supporting contributors, required evidence, SOP reference, instructions, risk and due-date conditions as applicable. Save a controlled draft, submit for independent review and follow the governed activation flow. SOP activity labels do not redefine protected lifecycle states or grant QA approval/signature powers.
What do QA and QC role labels mean?
QA and QC use uppercase abbreviations. Examples include QMS QA Manager, QMS QC Manager and QMS CAPA Owner. The display label is human-readable; the backend role or stage identifier and protected permissions do not change merely because a label is edited.
5. Updated homepage navigation and Ask Generics
The GenRx homepage now presents five product pathways, a searchable module directory with category selection, a dedicated product tour and a visible Ask Generics launcher/header entry. Public BayesPeaQ product information and the protected Quality workspace are distinct destinations. If Ask Generics fails to initialize, use the retry/reload guidance rather than treating an unanswered question as a scientific result.
Availability, permissions and specific control names can vary by release and tenant; the live authenticated workspace and its server-enforced permissions control what actions a user can perform.
BayesGenRx connects development questions to controlled evidence and accountable ANDA decisions.
Program OS is the operating record. Individual scientific, quality and regulatory workspaces should feed that record.
Select product, understand the RLD, choose a formulation, de-risk BE, control scale-up, answer a deficiency or lock a submission packet.
Keep official sources, measured development data, calculations, simulations and human review distinguishable and traceable.
Make formulation sensitivity, physiology, analytical limitations, source freshness and manufacturing risk visible.
Run another experiment, modify formulation, hold the stage gate, request review, prepare a study, or advance a controlled packet.
Four failure modes GenRx should prevent
Teams work from different RLD, labeling, Orange Book or PSG versions.
A simulation, prediction or inferred process is presented as measured or regulator-accepted evidence.
Formulation, BE, CMC, quality and submission decisions lose the common product identity and rationale.
A dashboard percentage appears green while a required source, review, measurement or gate remains unresolved.
The product identity, RLD basis, dosage form, strength, pathway, formulation revision, batches, evidence and review history should remain connected from first assessment through ANDA readiness.
Start with one real development question and keep the scope bounded.
A good first programme is a familiar oral solid dosage product with identifiable reference sources and a clear uncertainty.
- Create or open the programme in Program OS.
Confirm product, strength, dosage form, RLD/reference basis, market/pathway, owners and the current stage gate.
- Lock the official-source posture.
Review Orange Book, product labeling, PSG and relevant guidance. Record source identity and freshness before downstream conclusions depend on them.
- Characterize the reference product.
Capture observable physical, solid-state, formulation and performance evidence. Separate measured observations from reverse-engineering hypotheses.
- Define the QTPP and critical risks.
State the intended product profile and identify CQAs, material attributes and process variables likely to matter.
- Explore formulation and process options.
Compare composition, excipient roles, dissolution behavior, manufacturability and cost without losing revision history.
- Test product-performance similarity.
Use dissolution, biorelevant testing and IVIVC/PBBM context where appropriate. Interpret f2 and other metrics within their assumptions.
- Use GenRx vBE when the question is model-informative.
Compare Reference/Test product twins, virtual physiology, PK and BE uncertainty. Treat the output as development decision support.
- Plan or interpret BE evidence.
Connect study design, PK endpoints, statistics, subject variability and residual uncertainty to the programme.
- Control scale-up, stability and quality evidence.
Make process changes, transfer risks, deviations and supporting evidence explicit before packet lock.
- Build the ANDA story from the same record.
Use the ANDA Intelligence Board and regulatory workspaces to connect sources, Module 3 narrative, correspondence, eCTD readiness and reviewer signoff.
A formulation prediction, f2 result, PBPK/PBBM simulation or virtual BE result can be useful while still requiring additional experimental, quality, clinical or regulatory evidence.
Keep product identity, stage, evidence and ownership in one authoritative record.
Program OS is where the generic product journey is coordinated rather than a place to manually duplicate every scientific result.
| Programme element | What to record | Why it matters |
|---|---|---|
| Product identity | Drug, strength, dosage form and route. | Prevents evidence from being applied to the wrong presentation. |
| Reference basis | RLD/reference product, label/source identifiers and market. | Anchors sameness and comparative-development assumptions. |
| Pathway | ANDA posture and any unresolved pathway question. | Keeps development work aligned to the intended submission route. |
| Stage gate | Current development stage, blockers and next controlled action. | Prevents a dashboard score from silently authorizing progression. |
| Owners/reviewers | Accountable roles for formulation, BE, CMC, quality and regulatory work. | Separates work execution from approval authority. |
| Evidence spine | Linked source, experiment, analysis, model, review and packet objects. | Preserves why a decision was made and what can invalidate it later. |
Understand what the reference product tells you—and what it cannot reveal directly.
Reference-product characterization should distinguish observation, official information and development hypothesis.
Orange Book, labeling, PSG and guidance establish traceable regulatory context. Keep retrieval dates and versions.
Weight, dimensions, dissolution, solid state and analytical observations can support product understanding when methods and batches are known.
Likely excipient role, process or microstructure may guide development but should not be represented as known reference-manufacturing fact.
How should I interpret Q1 and Q2 information?
Q1 describes qualitative sameness of excipient identity; Q2 describes quantitative similarity. Their regulatory relevance depends on dosage form, pathway and product-specific expectations. GenRx should preserve the exact source or analytical basis for any Q1/Q2 conclusion and keep inferred composition separate from confirmed information.
Why can two visually similar tablets behave differently?
API solid state, particle size, excipient grade, granulation, compression, coating, porosity and microstructure can materially affect dissolution and absorption. Reference-product understanding therefore needs more than appearance or ingredient names.
Design against a declared target product profile, not by trial-and-error alone.
Use the QTPP to connect patient/regulatory expectations to CQAs, formulation choices, process variables and experiments.
Dosage form, strength, route, release behavior, stability, packaging and other product expectations.
Attributes such as assay, impurities, dissolution, content uniformity, hardness or moisture where relevant.
API attributes, excipient roles, grades, amounts and process steps that could affect CQAs.
Use experiments and modelling to understand sensitivity rather than treating a single formula as the whole knowledge base.
Preserve the rationale, trade-offs and unresolved risks when promoting a formula revision.
A recommended excipient level or process setting is useful only within the model, data and manufacturing context that produced it. It does not replace formulation scientist, CMC or quality review.
Use in-vitro performance to explain product behavior—not as a universal substitute for in-vivo evidence.
Dissolution becomes most useful when the method, medium, sampling, variability, formulation revision and intended decision are all explicit.
Inspect full profiles, variability and conditions—not only one summary value.
Interpret f2 within the applicable sampling and variability conditions. Do not generalize beyond the tested profiles.
Mechanistic or empirical models can support development questions when model qualification and context of use are explicit.
What if f2 is favorable but other evidence is not?
Keep the disagreement visible. A favorable f2 result does not erase differences in formulation, biorelevant behavior, variability, solid state or in-vivo exposure. The programme should identify what the metric answers and what it does not.
When should biorelevant dissolution be used?
Use it when physiologically relevant media or conditions can meaningfully test a performance hypothesis. It should be linked to the question being asked, not added merely because it is technically available.
Use virtual bioequivalence to test the development hypothesis before the pivotal study.
vBE combines product twins, dissolution/PBBM or PBPK anchoring, virtual physiology, PK, BE statistics, uncertainty and value-of-information analysis in one governed workspace.
Define the actual compared product states and the formulation/performance differences being tested.
Use measured or appropriately sourced product inputs. Keep assumptions visible where data are absent.
Propagate relevant physiology and variability rather than relying only on a typical subject.
Inspect AUC/Cmax Test/Reference behavior, intervals and simulated trial-level uncertainty.
Ask which product, model or physiological inputs drive the conclusion and whether the decision is fragile.
Prioritize the experiment, measurement or formulation change most likely to reduce decision-critical uncertainty.
Freeze the model version, inputs, assumptions, outputs, limitations and human review that support the next action.
GenRx vBE is development decision support. Whether modelling can replace, reduce or inform an in-vivo study depends on the product, evidence, applicable regulatory framework and authority review.
Connect study design, PK statistics and product evidence to one development conclusion.
A BE result should be interpreted together with formulation revision, batch identity, protocol, analytical method, subject variability and any special product-specific requirements.
| Area | Inspect | Decision boundary |
|---|---|---|
| Study design | Sequence, periods, washout, fasting/fed condition, replicate design where relevant. | Design suitability is a governed scientific/regulatory decision. |
| PK endpoints | AUC and Cmax definitions, sampling adequacy and analysis set. | Do not infer missing observed PK from simulation. |
| Statistics | Log-transformed treatment comparison, geometric mean ratios and confidence intervals as applicable. | Interpret within the relevant BE framework and product context. |
| Variability | Within-subject variability, outliers, dropouts and analytical performance. | High variability may change design strategy; it does not justify post-hoc relaxation. |
| Conclusion | Observed evidence, uncertainty, protocol deviations and reviewer rationale. | Final regulatory acceptability remains with qualified teams and authorities. |
Protect product performance while changing equipment, scale and operating conditions.
Scale-up evidence should connect material attributes and CPPs to CQAs, yield, dissolution and stability—not merely document that a batch was manufactured.
Is the intended equipment comparable in mechanism and controllable operating range?
Which CPP changes materially affect CQAs or product performance?
Are parameters, ranges, sampling, IPCs, acceptance criteria and responsibilities explicit?
Does the change require new comparability, validation, stability or regulatory evidence?
Use quality workflows to govern deviations, CAPA, evidence and accountable handoffs.
GenRx development context can inform quality work, but quality records have their own controlled roles, approvals, signatures and audit expectations.
Capture what happened, product/batch context, immediate containment and evidence.
Separate facts, hypotheses, evidence and reviewer disposition.
Assign accountable roles, due dates, verification and effectiveness review.
Keep who did what, when, why and under which authority reconstructable.
A formulation or BE conclusion does not automatically close a quality event. BayesPeaQ should preserve role-specific authority and audit history.
Build a reviewable submission story before the last mile.
The ANDA Intelligence Board links source posture, CMC narrative, evidence gaps, deficiency risk, correspondence and packet state to the same programme.
Track PSG, Orange Book, labeling and guidance movement before source changes invalidate a packet assumption.
Compose CMC narrative from controlled formulation, process, analytical, stability and quality evidence.
Link study and model evidence to the exact product/batch and conclusion it supports.
Check expected content, document state, version and reviewer posture before packet lock.
Map the question to evidence, owner, rationale, proposed response and unresolved risk.
Generate programme and board summaries without severing the underlying source/evidence links.
Later data can create a new revision, but should not silently rewrite what evidence existed when the earlier decision or packet was approved.
Do not collapse unlike evidence into a single readiness number.
Different evidence answers different questions. GenRx should preserve source, method, version, units, applicability, reviewer and unresolved uncertainty.
Traceable external source
Orange Book, PSG, labeling, guidance, literature or another controlled external source. Record identifier and retrieval date.
Calculation or model output
f2 calculation, PBPK/PBBM, vBE simulation, optimization or other algorithmic output with method/version and assumptions.
Observed development result
Analytical, dissolution, stability, PK or manufacturing measurement with batch, method, units and QC context.
Accountable interpretation
Formula promotion, risk acceptance, quality disposition, study interpretation or packet approval. Review is authority, not a measurement.
A disciplined decision workflow
- State the claim.
Example: “Formulation F07 is sufficiently de-risked to proceed to pilot BE.”
- Collect the relevant evidence without changing its class.
Official sources, dissolution, vBE, stability, manufacturing and prior study evidence can inform the same claim while remaining distinct.
- Check comparability.
Confirm product revision, strength, batch, method, condition, population/model context and date.
- Surface contradiction and missing evidence.
Do not let favorable simulation hide unfavorable measured data or stale official-source posture.
- Identify the decision-critical uncertainty.
Ask what could still change the next action and whether it can be reduced by the smallest useful experiment or review.
- Record the accountable decision.
Capture reviewer, rationale, conditions, limitations, evidence references and follow-up work.
Open the workspace that answers today's question.
Not every programme needs every module. Choose the smallest workflow that can resolve the current uncertainty and write the result back to the programme.
Server-controlled India/EU/US requirements, linked sources and multi-market evidence comparison.
Open →Governed modeling, QbD, stability, Source Watch and technical preflight decisions.
Open →Public illustrated journey; examples are not customer programme records.
Open →Programme master data, evidence, readiness, owners, gates and next actions.
Open →Source Watch, Module 3 composer, deficiency intelligence and packet export.
Open →Reference-product observations, composition/process hypotheses and development implications.
Open →API form, material evidence and solid-state development context.
Open →Formula candidates, excipient roles, trade-offs and development rationale.
Open →QTPP/CQA/CPP relationships and design-space exploration.
Open →Profile comparison and similarity calculations with explicit assumptions.
Open →Connect in-vitro performance to in-vivo behavior within a declared model context.
Open →Product twins, virtual physiology, PK/BE statistics, uncertainty and Decision Passport.
Open →Study evidence, PK/statistical interpretation and residual risk.
Open →Process sensitivity, equipment fit, transfer and validation readiness.
Open →Incidents, investigations, CAPA, role-aware quality work and audit trail.
Open →Controlled role workspaces, responsibility eligibility, SOP activities, review inbox and version history.
Open →Source posture, milestones, packet readiness and agency-facing evidence context.
Open →Submission structure and controlled dossier work.
Open →Connect filing windows, product economics and lifecycle decisions to the programme.
Open →Resolve the evidence or workflow reason for failure—do not force the programme green.
Missing, stale, contradictory, unavailable and review-required states are meaningful.
The programme shows a blocker even though a module has a result
Check whether the result is tied to the correct product revision, batch and programme, whether it has the required evidence class and review state, and whether another required domain is still missing. A computational result alone should not satisfy a measured or reviewed gate.
The official source changed after we started development
Do not overwrite the old source record. Register the new source/version, assess impact on assumptions and packet content, and record the resulting decision. Source Watch exists specifically to make this change visible.
vBE says the scenario is favorable but the BE study did not pass
Preserve both results. Investigate product/batch differences, model assumptions, physiology/variability, dissolution inputs, analytical/study execution and model applicability. Measured study evidence is not replaced by a favorable simulation.
f2 is unavailable or inappropriate
Keep the state explicit and use the scientifically appropriate alternative evidence or regulatory strategy. Do not manufacture an f2 number merely to complete a dashboard.
A quality task is assigned to the wrong person
Use BayesPeaQ's role-aware assignment/handoff workflow rather than changing the scientific programme owner. Quality responsibility and programme ownership are related but distinct authorities.
Scientific Requirements shows “Programme list unavailable”
First open Program OS and confirm an authorized programme exists and is explicitly selected. Anonymous Explore mode may prompt sign-in; expired credentials require a new session; a 403 means the tenant or user is not authorized. If the API is unavailable, retry when restored rather than treating an empty or failed list as valid source evidence.
My programme edits or evidence were not saved
Distinguish New programme from Save planning metadata. Verify the current version, the server receipt and subsequent read-back. A stale-version conflict, malformed response, missing product identity or unreviewed handoff should remain unresolved until the authoritative record confirms the change.
Why is my organization Quality role not selectable for an SOP stage?
Role Studio filters primary owners by approved stage eligibility. Check the active role catalogue, employee coverage and organizational assignment. Do not bypass the independent review or protected QA/QC responsibility boundary just to make a role selectable.
A service or model is unavailable
Preserve the unavailable/failed state and provenance. Retry when appropriate or select an explicitly supported alternative. Do not fabricate a scientific value or silently replace the intended method.
Two teams see different readiness
Verify they are viewing the same programme, product revision, source state and evidence revision. Readiness should be derived from authoritative programme evidence and gates, not browser-local assumptions.
Practical definitions for terms used across GenRx.
The important question is not only what the term means, but what it means for the decision.
| Term | Meaning in GenRx | Why it matters |
|---|---|---|
| ANDA | Abbreviated New Drug Application pathway for eligible generic drug products in the United States. | Defines the evidence and submission context for the programme. |
| RLD | Reference Listed Drug used as the regulatory reference in the U.S. context. | Reference identity must remain controlled across development. |
| PSG | Product-Specific Guidance published by FDA when available. | Can materially affect BE and development strategy. |
| QTPP | Quality Target Product Profile. | Connects intended product performance to development controls. |
| CQA | Critical Quality Attribute. | Attribute that should be controlled within an appropriate limit/range to ensure product quality. |
| CPP | Critical Process Parameter. | Process variable whose variation can affect CQAs. |
| Q1/Q2 | Qualitative excipient sameness / quantitative similarity terminology used in defined regulatory contexts. | Can shape formulation strategy but must be interpreted product-specifically. |
| f2 | Similarity factor used to compare dissolution profiles under applicable conditions. | A summary of profile similarity, not universal proof of in-vivo equivalence. |
| IVIVC | In-vitro/in-vivo correlation. | Links dissolution/release behavior to in-vivo performance within a qualified model context. |
| PBBM | Physiologically Based Biopharmaceutics Modelling. | Mechanistically connects product/formulation properties to absorption and exposure. |
| PBPK | Physiologically Based Pharmacokinetic modelling. | Represents drug disposition using physiology and drug-specific parameters. |
| vBE | Virtual bioequivalence simulation. | Tests a BE development hypothesis computationally; does not itself establish regulatory BE. |
| GMR | Geometric Mean Ratio, commonly Test/Reference for PK endpoints. | Core comparative statistic in BE assessment. |
| Decision Passport | Versioned record linking question, inputs, model/run, evidence, uncertainty, limitations and review. | Makes why a model-informed decision was made inspectable. |
| RETRIEVED | Evidence imported from a traceable external source. | Must retain source identity, version and relevance. |
| COMPUTED | Output generated by an algorithm, calculation or model. | Useful within declared assumptions and applicability. |
| MEASURED | Observed result tied to a method, sample/batch, units and QC. | Can satisfy experimental gates when identity and quality are established. |
| REVIEWED | Accountable human interpretation or governed decision. | Records authority and rationale; it is not itself a measurement. |
| Evidence spine | Connected history of source, experiment, analysis, model, review and packet objects for a programme. | Prevents context loss between functions and stages. |
| Stage gate | Declared condition that must be satisfied or reviewed before progression. | Prevents readiness scores from becoming automatic authority. |
Know what you own, what you contribute, and where independent review begins.
GenRx works best when teams share the same programme context but do not blur scientific, quality and regulatory authority.
Maintain the decision context
Confirm product identity, RLD basis, pathway, stage, owners, blockers and next decision. Coordinate work without converting dashboard readiness into automatic progression.
Program OS →Build and justify the product design
Own QTPP-to-formulation reasoning, excipient choices, QbD experiments, formula revisions and product-performance hypotheses.
Formulation Hub →Characterize comparative performance
Own method conditions, dissolution profiles, similarity interpretation, biorelevant evidence and model-ready product inputs.
Dissolution →Quantify exposure and uncertainty
Own vBE/PBBM/PBPK assumptions, virtual populations, PK endpoints, BE statistics, model limitations and study interpretation.
GenRx vBE →Protect performance through scale
Own material/process risk, scale-up assumptions, equipment fit, transfer controls, validation inputs and stability evidence.
Scale-Up Risk →Govern controlled quality decisions
Own quality events, investigation/CAPA governance, evidence integrity, independent review, signatures and closure according to policy.
BayesPeaQ →Keep the evidence aligned to the filing claim
Own official-source posture, pathway interpretation, dossier narrative, deficiency response and submission-state review.
ANDA Board →Do not hand off only a screenshot or a green status.
Example: move an immediate-release tablet from reference understanding to an ANDA-ready evidence story.
This is an operating example, not a regulatory conclusion for a specific product.
Create the programme with drug, strength, dosage form, intended U.S. ANDA pathway, RLD basis, owners and unresolved questions.
Output:Controlled programme context.Register current Orange Book, labeling, PSG/guidance and relevant official sources with retrieval dates.
Output:Traceable regulatory starting point.Measure and record reference-product physical, solid-state and dissolution evidence. Keep inferred process/composition as hypotheses.
Output:Reference evidence package.Define QTPP/CQAs, compare formulation candidates and test material/process risks with QbD evidence.
Output:Selected formulation revision + rationale.Compare dissolution profiles under declared conditions; use biorelevant or IVIVC/PBBM work when it answers a real uncertainty.
Output:Comparative performance assessment.Run Reference/Test product twins, virtual physiology, PK/BE simulations and uncertainty/VOI analysis using versioned inputs.
Output:Decision Passport + next experiment/study recommendation.Plan or interpret the appropriate BE study with exact formulation/batch identity, protocol, PK/statistics and deviations.
Output:Observed BE evidence and reviewed conclusion.Assess process/equipment changes, transfer, validation, stability and any deviations/CAPA without severing product revision identity.
Output:Controlled CMC/quality evidence.Connect Module 3, BE, source posture, quality evidence, correspondence and eCTD preflight to the same programme.
Output:Reviewable evidence-linked submission package.Use these as review prompts before major handoffs—not as automatic approval logic.
The applicable evidence depends on product, pathway, programme phase and organizational procedures.
- Correct RLD/reference basis and current source posture.
- Formula and manufacturing revision frozen for the intended study.
- Comparative dissolution and key product-performance risks reviewed.
- Critical model assumptions/uncertainty understood where vBE/PBBM informed the decision.
- Analytical, batch and protocol readiness reviewed by accountable functions.
- Open quality blockers explicitly dispositioned.
- Critical material attributes and CPP/CQA relationships understood.
- Equipment/mechanism differences assessed.
- Operating ranges, IPCs, sampling and acceptance criteria defined.
- Comparability, validation and stability obligations identified.
- Change-control/quality implications routed through BayesPeaQ where applicable.
- Official-source versions and product identity current.
- CMC narrative traces to controlled development/manufacturing evidence.
- BE/model evidence references exact formulation/batch and reviewed conclusions.
- Known contradictions, deficiencies and open commitments visible.
- Document versions, reviewers and approval state explicit.
- eCTD/preflight and accountable regulatory review completed.
Start in Program OS, then open only the workspace needed for the next development decision.
The goal is not to generate the most outputs. It is to keep the reference-to-ANDA evidence chain controlled, explainable and reviewable.