17 Feb 2022, 15:42
• The group discussed the value to testing AOPs by using novel compounds from chemical screening libraries like ToxCast that were previously shown to exhibit an activity associated with a KE
• Don't always see the predicted effects
• Noted the importance of reporting those results that don't agree with an AOP as well
• Ultimate goal is to improve the utility for application, not to "prove" the AOP
• Understanding why an AOP-based prediction didn't work provides an important means to refine and improve our understanding
• Also important to consider possible errors in the screening data - may not always mean the AOP was wrong - somethings there are different explanations
• Discussion about the best way to add discordant information to AOPs in the wiki
• Use of discussion pages, associated with each AOP page
• KEs and KERS can be edited by anyone with author rights - can add information to them, including information on inconsistencies and uncertainties
What topics of interest relative to AOPs haven't you heard so far in the CoP Symposium?
• Ways to deal with KEs that are synonyms or nearly synonyms
○ Key event components - partial solution, but variable use
○ Gardening - can help when issues are flagged, but labor intensive, and time-consuming
○ There has never been a systematic study of how good the key event components are at identifying synonymous events - that was their intent, but it has not been fully utilized for that purpose to date
• A number of the presentations talked about ascribing numerical indications of uncertainty to the KERs based on study quality, risk of bias, etc.
○ In practice, how is that actually done?
• Use of AOPs to support use of non-animal models
○ Build a case that animals aren't always the gold standard
○ In some cases, a non-animal in vitro assay etc. may be a better model for humans (or other animals of interest) than an intact animal model whose physiology is not conserved
How can AOP authors find out what AOPs are wanted/needed by regulators etc.?
• Would be nice to have a mechanism to link demand to development
• The AOP-Wiki seems like a good way to do it - apply a section to the AOP-Wiki or SAAOP
• NC3Rs could also be a repository for that
• Publish a putative AOPs - as crystallization points out into the community
• Researchers' desire to perform research that has impact is a key motivator for AOP development
• Would be really useful to researchers if they could identify a wish list of AOPs that regulators would like to see developed or improved, that would be really valuable
• Regulators may not know what they need yet, because not a ton of application examples yet
• Would be valuable to have a place where they could add ideas, in the moment as they occur
• A list of AOPs of interest would also be really helpful for students - educational programs related to AOP development could develop student projects related to developing AOPs that have been requested/are desired.
• Don't always see the predicted effects
• Noted the importance of reporting those results that don't agree with an AOP as well
• Ultimate goal is to improve the utility for application, not to "prove" the AOP
• Understanding why an AOP-based prediction didn't work provides an important means to refine and improve our understanding
• Also important to consider possible errors in the screening data - may not always mean the AOP was wrong - somethings there are different explanations
• Discussion about the best way to add discordant information to AOPs in the wiki
• Use of discussion pages, associated with each AOP page
• KEs and KERS can be edited by anyone with author rights - can add information to them, including information on inconsistencies and uncertainties
What topics of interest relative to AOPs haven't you heard so far in the CoP Symposium?
• Ways to deal with KEs that are synonyms or nearly synonyms
○ Key event components - partial solution, but variable use
○ Gardening - can help when issues are flagged, but labor intensive, and time-consuming
○ There has never been a systematic study of how good the key event components are at identifying synonymous events - that was their intent, but it has not been fully utilized for that purpose to date
• A number of the presentations talked about ascribing numerical indications of uncertainty to the KERs based on study quality, risk of bias, etc.
○ In practice, how is that actually done?
• Use of AOPs to support use of non-animal models
○ Build a case that animals aren't always the gold standard
○ In some cases, a non-animal in vitro assay etc. may be a better model for humans (or other animals of interest) than an intact animal model whose physiology is not conserved
How can AOP authors find out what AOPs are wanted/needed by regulators etc.?
• Would be nice to have a mechanism to link demand to development
• The AOP-Wiki seems like a good way to do it - apply a section to the AOP-Wiki or SAAOP
• NC3Rs could also be a repository for that
• Publish a putative AOPs - as crystallization points out into the community
• Researchers' desire to perform research that has impact is a key motivator for AOP development
• Would be really useful to researchers if they could identify a wish list of AOPs that regulators would like to see developed or improved, that would be really valuable
• Regulators may not know what they need yet, because not a ton of application examples yet
• Would be valuable to have a place where they could add ideas, in the moment as they occur
• A list of AOPs of interest would also be really helpful for students - educational programs related to AOP development could develop student projects related to developing AOPs that have been requested/are desired.