Help: Getting Started

The AOP Framework

AOP Principles

Adverse Outcome Pathways (AOPs) are a conceptual framework for organizing, synthesizing, and presenting specialized scientific knowledge regarding the linkage between perturbation of a specific biological target, pathway or process by a stressor (e.g., a chemical, radiation, a virus), and a consequent adverse outcome(s) considered relevant to risk assessment, regulatory decision-making, and/or environmental management. They are intended to aggregate knowledge currently dispersed in journal articles, textbooks, reports, databases, and many other sources into a systematic and accessible format that facilitates use of that knowledge to support a broad diversity of applications and interests.

AOPs are described according to five key principles.

1.    AOPs are not stressor specific – They describe generalized motifs of biological response to a specific type of biological perturbation that apply without regard to which specific stressor initiated the response(s).

2.    AOPs are modular – AOPs are described using two basic units of information:  (1) key events that describe measurable biological responses with some understood causal relationship to downstream biological change; and (2) key event relationships that lay out the evidence supporting a causal relationship between any pair of key events (one causing the other) and our understanding under which event A can cause event B to occur. Those units may be shared among multiple AOPs.

3.    AOPs are a pragmatic unit of development and review – An individual AOP defines a single sequence of biological events leading from a single initiating event to a single adverse outcome. It is recognized that biological response to stressors is not that simplistic. However, an AOP does not try to capture ALL possible outcomes or pathways to those outcomes, just a single pathway from initiation to outcome. This is considered a tractable unit for development and review.

4.    Modular AOPs can assemble into AOP networks – An AOP network is a group of two or more AOPs that share one or more key events. Assembly of individual AOPs into AOP networks provides the ability to capture the greater complexity of biological responses in an organized and systematic way.

5.    AOPs are living documents – Because AOP descriptions are a synthesis and integration of existing knowledge, they reflect the current state of science and knowledge and the AOP authors expertise and investment in organizing and communicating the current state. AOPs with knowledge gaps and uncertainties can still be useful, and even after publication, AOPs can continue to evolve as new data and knowledge are generated. Thus, AOPs should be evaluated based on the extent to which they capture current understanding and effectively communicate it to non-specialist audiences that need to use it, not whether they can address all questions concerning the impact of a given biological perturbation.

AOP descriptions developed in the AOP-Wiki are organized into three distinct page types. Each page type includes a template that requests specific type of information.

Event Pages

Event pages describe “Key Events” in an AOP. A key event is a measurable change in a biological system (i.e., compared to a control). It is considered essential to, but not necessarily sufficient for progression to an adverse outcome. Each key event description should include three critical pieces of information.

Table 2.  Key information to include in a key event description (Event page).

Information Field Description
Key Event Description The author(s) should clearly describe what it is (process, endpoint) that can be measured.
How it is measured/detected The author(s) should cite and describe methods that are used to make the measurement associated with the key event. If the measurement is part of a standardized test guideline or established commercial or programmatic assay, the corresponding guideline or assay should be cited.
Domain of Applicability The authors(s) should define the life stages, sex(es), and taxonomic groups for which making the measurement is relevant. The authors should also provide their rationale and the underlying scientific support for that rationale.

Relationship Pages

Relationship pages provide the scientific support for concluding that two key events (Event A and Event B) are causally-related to one another (i.e., Event A leads to or can cause Event B).

  • Each Key Event Relationship description should briefly describe the two events being linked together, and their relation to one another (i.e., which is upstream and which is downstream).
  • Evidence Supporting this KER encapsulates the scientific evidence/support for inferring that if Event A is impacted, some effect on Event B could reasonably be expected.

Line of Evidence


Biological Plausibility

What is known about the relationship between Event A and Event B based on fundamental understanding of normal biological function (i.e., in the absence of stressor). Structural or functional relationships between Event A and Event B.

Empirical support

  • Temporal concordance

Evidence that upon exposure to a stressor, Event A is impacted before (earlier in a time-course) than Event B.

Empirical support

  • Dose concordance

Evidence that upon exposure to a stressor, Event A is impacted at a magnitude of exposure that is less than or equal to that which is shown to impact Event B.

Empirical support

  • Incidence concordance

Evidence that upon exposure to equivalent magnitudes of stressor, the proportion of the population showing a significant response associated with Event A is greater than or equal to the proportion of the population showing a significant response associated with Event B.

Essentiality (optional for KER)

Evidence that upon exposure to a stressor, if Event A is prevented from occurring (e.g., through knock-out, over-expression, pharmaceutical manipulation, etc.), Event B is also prevented.

Uncertainties or Inconsistencies

The author should report any evidence that appears contradictory to a presumed causal relationship between Event A and Event B, or gaps in the understanding of why a causal relationship would between Event A and Event B would exist.

  • In addition to evidence supporting a causal relationship between Event A and Event B, authors are also encouraged to provide Quantitative Understanding of the Linkage. Quantitative understanding refers to information that helps to define how much change in Event A, and under what conditions, is required to elicit some magnitude of change in Event B. Where available, authors are asked to assemble four types of information about the quantitative understanding of the linkage.

Desired Information


Response-response relationships

Information concerning the type of mathematical function, plot, or model describes the magnitude of change in Event B (dependent variabile) as a function of the magnitude/duration of effect on Event A (independent variable).

Time scale

Information that helps to define the relative time scale over which Event B is expected to respond following some effect on Event A.  Seconds, minutes, hours, days, weeks, ……

Known modulating factors

Information concerning variables that are known to alter or shift the response-response relationship between Event A and Event B. Examples include genetic predispositions, diet, pre-existing disease states, co-occurring stressors, social stress, etc.

Known positive or negative feedback loops

Information concerning ways that changes in Event B can in turn influence the state of Event A. It can be expected that in many cases, Event A is not independent of Event B, even though it may initially be the causal part of the pairing.

AOP Pages

Adverse outcome pathway pages summarize a sequence of key events connecting a molecular initiating event to an adverse outcome. The key events and relationships included in the AOP are listed in table format and as a graphical box and arrow diagram. The major role of the AOP page is to provide an overall assessment of the AOP, considering weight of evidence across the span of Key Event Relationships included in the AOP.

Domain of applicability – as for the Event and Relationship pages, the domain of applicability identifies the sexes, life stages, and taxon for which the AOP is expected to be relevant and provides the rationale for that expectation. The key difference, is that the domain of applicability is now considered across the entire span of the AOP.  

Essentiality of the Key Events – Key event relationship descriptions consider evidence that upon exposure to a stressor, if Event A is prevented (e.g.,, through knock-out, over-expression, pharmaceutical manipulation, etc.) from occurring, Event B is also prevented. On the AOP page, consideration of essentiality of the key events is extended considering the potential effects of modulation or prevention of Key Event A on all subsequent key events in the sequence. The more key events along the AOP for which there is compelling evidence that they play an essential causal role in the progression of the toxic effect, the greater the overall confidence in the AOP.

Evidence Assessment - The aim of the evidence assessment on the AOP page is not to restate all the evidence and information captured on the individual Relationship pages, but to provide an overall perspective on the strengths and limitations of the AOP and the rationale for identifying those. Weight of evidence “calls” are made by the authors based on a series of guiding questions from the Users’ Handbook. The aim on the AOP page is to provide the overall rationale and justification for the weight of evidence “calls” that were assigned to the key event relationships, then discuss how the overall level of support for the AOP may impact its fit-for-purpose for various applications. This section can also be used to identify critical data gaps which, if addressed, could further enhance the utility of the AOP.

Considerations for Potential Applications of the AOP – In light of the overall assessment of the AOP (domain of applicability, evidence for essentiality, evidence assessment) authors are invited to discuss some of the potential applications of the AOP. However, it is recognized, that there can be a wide range of possible applications, and it might be difficult for any set of authors to be comprehensive.