NTR: New histone reader activities

[pgaudet]

For the work on histone binding/histone readers here:
geneontology/go-annotation#5203

@colinlog and I have created the following terms

+[Term]
+id: GO:0140063
+name: unmodified histone reader activity
+namespace: molecular_function
+def: "A histone reader that specifically binds either to an unmodified histone." [GOC:pg]
+comment: Curator note: To annotate the specific histone recognized, use 'has input' extension.
+is_a: GO:0140566 ! histone reader activity
+relationship: has_part GO:0035064 ! methylated histone binding
+property_value: term_tracker_item "#29384"
xsd:anyURI
+created_by: pg
+creation_date: 2024-12-13T14:55:58Z

+[Term]
+id: GO:0140054
+name: histone H2A reader activity
+namespace: molecular_function
+def: "A histone reader that specifically binds either to an unmodified histone H2A or a form
modified by a post-translational modification on a specific residue. The most common PTMs on h
istones are methylation, acetylation and phosphorylation." [PMID:11498575, PMID:25688442, PMID
:31082667, PMID:34726351]
+synonym: "histone H2A reader" EXACT []
+is_a: GO:0140566 ! histone reader activity
+relationship: has_part GO:0042393 ! histone binding
+relationship: occurs_in GO:0000786 ! nucleosome
+created_by: pg
+creation_date: 2024-12-13T14:14:39Z

+[Term]
+id: GO:0140071
+name: histone H2B reader activity
+namespace: molecular_function
+def: "A histone reader that specifically binds either to an unmodified histone H2B or a form
modified by a post-translational modification on a specific residue. The most common PTMs on h
istones are methylation, acetylation and phosphorylation." [PMID:11498575, PMID:25688442, PMID:31082667, PMID:34726351]
+synonym: "histone H2A reader" EXACT []
+is_a: GO:0140566 ! histone reader activity
+relationship: has_part GO:0042393 ! histone binding
+relationship: occurs_in GO:0000786 ! nucleosome
+created_by: pg
+creation_date: 2024-12-13T14:57:55Z
+
+[Term]
+id: GO:0140072
+name: histone H3K9ac reader activity
+namespace: molecular_function
+def: "A histone reader that recognizes a histone H3 acetylated at lysine 9." [GOC:pg]
+synonym: "H3K9ac modified histone binding" NARROW []
+is_a: GO:0140006 ! histone H3 reader activity
+relationship: has_part GO:0035064 ! methylated histone binding
+property_value: term_tracker_item "#29384" xsd:anyURI
+created_by: pg
+creation_date: 2024-12-13T14:59:50Z

+[Term]
+id: GO:0140055
+name: histone H4K8ac reader activity
+namespace: molecular_function
+def: "A histone reader that recognizes a histone H4 acetylated at lysine 8." [GOC:pg]
+synonym: "H4K8ac modified histone binding" NARROW []
+is_a: GO:0140008 ! histone H4 reader activity
+relationship: has_part GO:0035064 ! methylated histone binding
+property_value: term_tracker_item "#29384"
xsd:anyURI
+created_by: pg
+creation_date: 2024-12-13T14:54:04Z

Colin will provide references for terms that don't have yet.

[ValWood]

Why not do these with PRO, since that is what we have already done with adaptors in the causal direction?

1. The type of modification doesn't provide the specificity we need for describing the code (need residue level)
2. It seems odd that the "causal" direction (i.e. recruitment based on a prior upstream modification) has less specificity (ie. we would not be allowed to have histone H3-Me adaptor activity (and this would not be specific enough on its own, we would still need the PRO ID to describe it fully)
3. we would need a huge number of combinations to describe these.

e.g:

htb1 https://www.pombase.org/gene/SPCC622.09

chromatin-protein adaptor activity
active form htb1 (MonoUbiq(K119)) recruits wdr70 part of double-strand break repair via homologous recombination

hht3 https://www.pombase.org/gene/SPBC1105.11c
chromatin-protein adaptor activity
active form hht3 (Phos(T3)) recruits bir1 part of mitotic sister chromatid biorientation
active form hht3 (Me(K9)) recruits clr4, swi6 part of pericentric heterochromatin formation
active form hht3 (Me(K9),Phos(S10)) recruits chp1, swi6 part of heterochromatin formation

rpb1 rpb1 https://www.pombase.org/gene/SPBC28F2.12
chromatin-protein adaptor activity
active form rpb1 (Phos(CTD-S5)) recruits cdk9, pcm1, set1 part of transcription elongation-coupled chromatin remodeling
active form rpb1 (Phos(CTD-S5)) recruits ceg1, pct1 part of 7-methylguanosine mRNA capping in chromatin
active form rpb1 (Phos(CTD-S2)) recruits rhn1
active form rpb1 (Phos(CTD-S2)) recruits seb1 part of termination of RNA polymerase II transcription
active form rpb1 (Phos(CTD-S2)) recruits set2 part of transcription elongation-coupled chromatin remodeling
active form rpb1 (Phos(CTD-S7)) recruits cdk9
active form rpb1 (Phos(CTD-S2,S5,S7)) recruits pin1 part of positive regulation of transcription initiation by RNA polymerase II during cellular response to oxidative stress
active form rpb1 (Phos(CTD-T4)) part of co-transcriptional lncRNA 3' end processing, cleavage and polyadenylation pathway part of intracellular phosphate ion homeostasis
active form rpb1 (Phos(CTD-T4)) part of positive regulation of transcription initiation by RNA polymerase II part of intracellular phosphate ion homeostasis
active form rpb1 (Phos(CTD-S5,S7)) part of intracellular phosphate ion homeostasis
active form rpb1 (Phos(CTD-S2,S5)) recruits ceg1 part of 7-methylguanosine mRNA capping in chromatin

[pgaudet]

Hi @ValWood

Why not do these with PRO,

You mean as annotation extensions?

[ValWood]

For the causal direction (see the linked examples above) the PRO ID is the annotated object.

We have not curated many "readers" because we curated them in the causal direction which makes more sense for GO cams
(i.e signalling pathway activates 'modifier'; modifier modifies histone; histine recruit 'reader'; 'reader is part of process x.

but for those we have annotated, we have done in the same way:
https://www.pombase.org/term/GO:0140566
(note that the ubiquitin-modified histone reader activity is only under binding).

We need the residue detail , so we need to use PRO anyway.
Without the residue detail, we can't disambiguate (the rpb1 is the best example of that because we have curated more comprehensively).

[ValWood]

actually for the readers the PRO is the extension.

in the causal direction the PRO is the annotated object.

Curating the readers is a bit of a duplication of effort because we have to do them in both directions.

[ValWood]

See comment here
#23932 (comment)

There is also an added level of complication if you go the specific GO term per modification route.
What do you do in cases where multip[le types of modification are required?

For example

active form hht1 (Me(K9),Phos(S10)) recruits chp1, swi6 part of heterochromatin formation

active form rpb1 (Phos(CTD-S2,S5,S7)) recruits pin1 part of positive regulation of transcription initiation by RNA polymerase II during cellular response to oxidative stress

active form rpb1 (Phos(CTD-S2,S5)) recruits ceg1 part of 7-methylguanosine mRNA capping in chromatin

[ValWood]

We don't know the precise residue codes yet, but we want to capture them when known. The best example above is that
chp1 and swi6 'read' histone hht1 with Me-K9 AND P-S10 simultaneously.

[pgaudet]

Hi @ValWood

I dont know what relation 'recruits' represents, so it's hard to comment on this.

I agree PRO is great when it's available, but I dont think every organism has PRO forms, or has the ability to use them.

@colinlog and I are looking for a compromise that allows to capture the main steps. Certainly looking at GO-CAMs would be very informative.

Thanks, Pascale

[pgaudet]

Colin confirms that the general references for H2A and H2B readers are correct.

Added references:
GO:0140055 histone H4K8ac reader activity >>> PMID:17996705
GO:0140072 histone H3K9ac reader activity >>> PMID:26067602

[pgaudet]

@ValWood

(note that the ubiquitin-modified histone reader activity is only under binding).

I dont see that

[ValWood]

Re "I dont see that"

I see this:

Some readers appear under MF and some under protein binding. I think it may have something to do with has part?

[ValWood]

Got it.
Histone reader isn't under protein binding
ubiquitin-modified protein reader activity is under protein binding
(I think possibly the relationship to protein binding in these cases should be has_part)

[pgaudet]

ubiquitin-modified protein reader activity

OK

I was looking at ubiquitin-modified histone reader activity
I haven't fixed all the modified binding terms (I first need to check annotations).

[pgaudet]

For ubiquitin-modified protein reader activity I changed the relation to 'has part' 'modification-dependent protein binding'

but it was already is_a protein-macromolecule adaptor activity