ChEMBL evidence strings: up_or_down activity for drugs

[deniseOme]

Following a support email "I totally understand what "antagonist" and "agonist" mean in the "activity" column. However, I am not able to uncover the meaning of "up_or_down". Could you help me on that? Where could I find this information?"

According to @gkos-bio "It means mainly that this was a temporary solution until we could better qualify the MOA of the small molecule or biologic.
Maybe, it’s time to revisit the classification of MOAs in Open Targets. This will require more work from and coordination with ChEMBL."

From Paula (ChEMBL):
_up_or_down seems to be one of the activity types in the JSON schema:

"activity": {
"type": "string",
"enum": [
"http://identifiers.org/cttv.activity/decreased_transcript_level",
"http://identifiers.org/cttv.activity/decreased_translational_product_level",
"http://identifiers.org/cttv.activity/drug_negative_modulator",
"http://identifiers.org/cttv.activity/drug_positive_modulator",
"http://identifiers.org/cttv.activity/gain_of_function",
"http://identifiers.org/cttv.activity/increased_transcript_level",
"http://identifiers.org/cttv.activity/increased_translational_product_level",
"http://identifiers.org/cttv.activity/loss_of_function",
"http://identifiers.org/cttv.activity/partial_loss_of_function",
"http://identifiers.org/cttv.activity/up_or_down",
"http://identifiers.org/cttv.activity/up",
"http://identifiers.org/cttv.activity/down",
"http://identifiers.org/cttv.activity/tolerated",
"http://identifiers.org/cttv.activity/predicted",
"http://identifiers.org/cttv.activity/damaging",
"http://identifiers.org/cttv.activity/damaging_to_target",
"http://identifiers.org/cttv.activity/predicted_tolerated",
"http://identifiers.org/cttv.activity/predicted_damaging",
"http://identifiers.org/cttv.activity/tolerated_by_target",
"http://identifiers.org/cttv.activity/unknown"
],
"description": "Activity of target in disease context"
In our data we assign 3 types of activities:

drug_negative_modulator for drugs which are negative modulators,
drug_positive_modulator for drugs which are positive modulators,
up_or_down for any other

So I think this must have been how it was required (from many years ago I think)._

This is how ChEMBL classifies the different activity types (what they seem to provide to us is
the parent term i.e. negative modulator, positive modulator. But do we have "other". Or do we use "up_or_down" instead of "other"):

NEGATIVE MODULATOR (parent type)	BLOCKER (action type)	Negatively effects the normal functioning of an ion channel e.g., prevents or reduces transport of ions through the channel (description)
NEGATIVE MODULATOR	RELEASING AGENT	Reverses the normal functioning of a transporter, causing release of the substrate, rather than uptake
NEGATIVE MODULATOR	ANTAGONIST	Binds to a receptor and prevents activation by an agonist through competing for the binding site
NEGATIVE MODULATOR	INVERSE AGONIST	Binds to and inactivates a receptor
NEGATIVE MODULATOR	ANTISENSE INHIBITOR	Prevents translation of a complementary mRNA sequence through binding to it
NEGATIVE MODULATOR	ALLOSTERIC ANTAGONIST	Binds to a receptor at an allosteric site and prevents activation by a positive allosteric modulator at that site
NEGATIVE MODULATOR	NEGATIVE ALLOSTERIC MODULATOR	Reduces or prevents the action of the endogenous ligand of a receptor through binding to a site distinct from that ligand (non-competitive inhibition)
NEGATIVE MODULATOR	INHIBITOR	Negatively effects (inhibits) the normal functioning of the protein e.g., prevention of enzymatic reaction or activation of downstream pathway
NEGATIVE MODULATOR	DEGRADER	Binds to or antagonizes a receptor, leading to its degradation
NEGATIVE MODULATOR	NEGATIVE MODULATOR	Negatively effects the normal functioning of a protein e.g., receptor antagonist, inverse agonist or negative allosteric modulator
OTHER	SUBSTRATE	Carried by a transporter, possibly competing with the natural substrate for transport
OTHER	MODULATOR	Effects the normal functioning of a protein in some way e.g., mixed agonist/antagonist or unclear whether action is positive or negative
OTHER	CHELATING AGENT	Binds to a metal, reducing its availability for further interactions
OTHER	HYDROLYTIC ENZYME	Hydrolyses a substrate through enzymatic reaction
OTHER	DISRUPTING AGENT	Destabilises or disrupts a protein complex, macromolecular assembly, cell membrane etc
OTHER	SEQUESTERING AGENT	Binds to a substance such as a drug, toxin or metabolite reducing its availability for further interactions
OTHER	METHYLATING AGENT	Methylates or participates in methylation (e.g., through donation of a methyl group) of a substrate molecule
OTHER	STABILISER	Increases the conformational stability of a protein or complex
OTHER	REDUCING AGENT	Modifies a substrate via a reduction reaction
OTHER	OTHER	Other action type, not clearly postively or negatively affecting the normal functioning of a protein e.g., chelation of substances, hydrolysis of substrate
OTHER	PROTEOLYTIC ENZYME	Hydrolyses a protein substrate through enzymatic reaction
OTHER	BINDING AGENT	Binds to a substance such as a cell surface antigen, targetting a drug to that location, but not necessarily affecting the functioning of the substance itself
OTHER	OXIDATIVE ENZYME	Oxidises a substrate through enzymatic reaction
OTHER	CROSS-LINKING AGENT	Induces cross-linking of proteins or nucleic acids
POSITIVE MODULATOR	OPENER	Positively effects the normal functioning of an ion channel e.g., facilitates transport of ions through the channel
POSITIVE MODULATOR	POSITIVE MODULATOR	Positively effects the normal functioning of a protein e.g., receptor agonist, positive allosteric modulator, ion channel activator
POSITIVE MODULATOR	ACTIVATOR	Positively effects the normal functioning of the protein e.g., activation of an enzyme or cleaving a clotting protein precursor
POSITIVE MODULATOR	AGONIST	Binds to and activates a receptor, often mimicking the effect of the endogenous ligand
POSITIVE MODULATOR	POSITIVE ALLOSTERIC MODULATOR	Enhances the action of the endogenous ligand of a receptor through binding to a site distinct from that ligand
POSITIVE MODULATOR	PARTIAL AGONIST	Binds to and only partially activates a receptor (relative to the response to a full agonist)

Things we may need to do:

1. revisit our classification of activity

• Do we want to change it to be in line with ChEMBL's (table above)?
• If we continue with "agonist", "antagonist" and "up_or_down", we need a sentence in the external docs (FAQ or other) to explain what it is.

Considerations:
Where is this activity data in the ChEMBL website? Is it under "activity charts"?
If so, there does not seem to be a correspondence between what we have ("up_or_down" for example) and what they provide. Ex:

• CHEMBL2108576
  --> Platform: up_or_down in the drug table (https://www.targetvalidation.org/evidence/ENSG00000142192/EFO_0000249?view=sec:known_drug)

--> ChEMBL website: no data available: https://www.ebi.ac.uk/chembl/compound_report_card/CHEMBL2108576/

• CHEMBL1201550
  --> Platform: up_or_down in the drug table (https://www.targetvalidation.org/evidence/ENSG00000134460/EFO_1000785?view=sec:known_drug)

--> ChEMBL website: other, plus hepatotoxicity
https://www.ebi.ac.uk/chembl/compound_report_card/CHEMBL1201550/
(if we click on "other" in the bioactivity summary chart we get redirected to: https://www.ebi.ac.uk/chembl/g/#browse/activities/filter/molecule_chembl_id%3A(%22CHEMBL1201550%22)%20AND%20standard_type%3A(%22Hepatotoxicity%20(cytolytic)%22%20OR%20%22Hepatotoxicity%20(granulomatous%20hepatitis)%22%20OR%20%22Hepatotoxicity%20(malignant%20tumour)%22%20OR%20%22Hepatotoxicity%20(mechanism)%22%20OR%20%22Hepatotoxicity%20(severe%20hepatitis)%22%20OR%20%22Hepatotoxicity%20(steatosis)%22%20OR%20%22Hepatotoxicity%20(successful%20reintroduction)%22%20OR%20%22Hepatotoxicity%20(time%20to%20onset)%22)

• CHEMBL1655
  --> Platform: up_or_down in the drug table
  https://www.targetvalidation.org/evidence/ENSG00000091831/EFO_0000616?view=sec:known_drug (modulator in the drug summary page: https://www.targetvalidation.org/summary?drug=CHEMBL1655)

ChEMBL website: Inhibition...
https://www.ebi.ac.uk/chembl/compound_report_card/CHEMBL1655/

• CHEMBL1201649
  --> Platform: agonist in the drug table (https://www.targetvalidation.org/evidence/ENSG00000091831/EFO_0000616?view=sec:known_drug)

--> ChEMBL website: no data available:
https://www.ebi.ac.uk/chembl/compound_report_card/CHEMBL1201649/

@AsierGonzalez says "I guess it would be interesting to know how many drugs there are for each activity category. Perhaps there are only a handful of “up_or_down” cases.

[AsierGonzalez]

I have checked the numbers of each of the different categories in the evidence data from the 19.02 release and this is what I found:

Counts based on the target.activity field in the JSON

Target activity	Counts
drug_negative_modulator	391178
drug_positive_modulator	63493
up_or_down	22293

Counts based on the action_type field in the JSON

Target activity	Action type	Counts
NEGATIVE MODULATOR	RELEASING AGENT	340
NEGATIVE MODULATOR	ANTAGONIST	27840
NEGATIVE MODULATOR	INVERSE AGONIST	347
NEGATIVE MODULATOR	ANTISENSE INHIBITOR	0
NEGATIVE MODULATOR	ALLOSTERIC ANTAGONIST	192
NEGATIVE MODULATOR	NEGATIVE ALLOSTERIC MODULATOR	3310
NEGATIVE MODULATOR	INHIBITOR	332247
NEGATIVE MODULATOR	DEGRADER	1
NEGATIVE MODULATOR	NEGATIVE MODULATOR	10
NEGATIVE MODULATOR	BLOCKER	10
OTHER	SUBSTRATE	20
OTHER	MODULATOR	18635
OTHER	CHELATING AGENT	0
OTHER	HYDROLYTIC ENZYME	1521
OTHER	DISRUPTING AGENT	135
OTHER	SEQUESTERING AGENT	0
OTHER	METHYLATING AGENT	0
OTHER	STABILISER	750
OTHER	REDUCING AGENT	0
OTHER	OTHER	340
OTHER	PROTEOLYTIC ENZYME	16
OTHER	BINDING AGENT	538
OTHER	OXIDATIVE ENZYME	0
OTHER	CROSS-LINKING AGENT	336
POSITIVE MODULATOR	OPENER	1701
POSITIVE MODULATOR	POSITIVE MODULATOR	4816
POSITIVE MODULATOR	ACTIVATOR	2108
POSITIVE MODULATOR	AGONIST	35341
POSITIVE MODULATOR	POSITIVE ALLOSTERIC MODULATOR	18135
POSITIVE MODULATOR	PARTIAL AGONIST	1392

For me the biggest question is why we seem to use "antagonist", "agonist" and "up_or_down" as activities in the platform, which is a mixture of the activity and action data we get from ChEMBL.

[MichaelaEBI]

I will investigate this with Paula from ChEMBL as part of an effort to understand all fields in the ChEMBL evidence strings.

[MichaelaEBI]

Update: It looks like ChEMBL does the mapping of different activity types to drug_positive_modulator, drug_negative_modulator and up_or_down because they provide these terms in the target.activity field and they also provide their activity types in the target2drug.action_type field.

The target.activity field entries are changed to agonist, antagonist and up_or_down in the front end.

It makes sense for us to use the terms in the action_type field to be more specific. This information has been added to this document which specifies upcoming changes to the JSON schema for ChEMBL evidence strings.

[MichaelaEBI]

For the 19.09 release, ChEMBL have changed the activity values to be negative_modulator, positive_modulator and other. The plan is to display both, these general categories and ChEMBL's more specific terms in the drug tables.

[MichaelaEBI]

This ticket can be closed since details for the front end have been summarised in ticket #730.