Menu
▾
▴
#782 New term request: Chromatin modifier-associated region identification by ChIP-Seq assay
I'm trying to get new terms to describe some of the data in the FaceBase resource (this project is funded by NIH/NIDCR, website at facebase.org).
- Suggested term name: Chromatin modifier-associated region identification by ChIP-Seq assay. There is currently a term for "transcription factor binding site identification by ChIP-Seq assay" so this could be added as a sibling. This is meant to cover chromatin modifiers like EP300, which is a histone acetyltransferase. So it is not a histone modification, but is the enzyme that adds a histone modification. It can refer to GO:0016568
- Synonyms: chromatin modifier binding site identification by ChIP-Seq assay
- Definition: A ChIP-seq assay to identify regions of chromatin bound to a chromatin modifying protein, e.g. a histone acetyltransferase.
- Definition source: PMID 19212405
- Parent term: suggest new term "reporter gene assay" (Definition: An assay in which expression of a reporter gene is detected that was inserted under the control of a regulatory sequence of interest.), with the existing term "Promoter activity detection by reporter gene assay" as a child of this same parent.
- Example of usage: PMID 19212405: "we present the results of chromatin immunoprecipitation with the enhancer-associated protein p300 followed by massively parallel sequencing, and map several thousand in vivo binding sites of p300 in mouse embryonic forebrain, midbrain and limb tissue"
- Editors: Paul D. Thomas, Yang Chai
Thanks for submitting. We plan to start working on them and hope to be back in touch soon. If you are available for the Oct. 12 obi-dev call we can discuss then.
EP300 is also a transcription factor (GO:0001102 as listed on http://www.ncbi.nlm.nih.gov/gene/328572.
Are there chromatin modifiers that are not also considered transcription factors? If not it's not clear that this would be a sibling term.
An example of a chromatin modifier that is not a transcription factor would
be sometime that modifies histones such as a histone deacetylase.
On Mon, Oct 12, 2015 at 12:25 PM Chris Stoeckert stoeckert@users.sf.net
wrote:
Related
OBI Terms:
#782Sorry but histone deacetylase is a functional class and not an example. Hdac1 is a histone deacetylase that is also annotated as a transcription factor (http://www.ncbi.nlm.nih.gov/gene/433759). I guess I should have just said that my understanding is that all chromatin modifiers that one would ChIP-seq to find their binding site are transcription factors.
The ENCODE listing at UCSC has a matrix of ChIP-seq with just two categories: Histone Modification and Transcription Factor (https://genome.ucsc.edu/ENCODE/dataMatrix/encodeChipMatrixHuman.html) where EP300 and HDAC1 are listed under Transcription Factor. To me this reflects the common usage and expectation for where to find these chromatin modifiers. If FaceBase needs a further specification of ChIP-seq of chromatin modifiers then we should have a subtype of "transcription factor binding site identification by ChIP-Seq assay” rather than a sibling class to avoid EP300 ChiP-seq being inconsistently assigned to one or the other.
Chris
Related
OBI Terms:
#782Just to be clear, no intent to argue other than to undersand what the
answer to issue might be, which I found intriguing, particularly after your
response, Chris.
On Mon, Oct 12, 2015 at 4:04 PM Chris Stoeckert stoeckert@users.sf.net
wrote:
Identifying Differential Histone Modification Sites from ChIP‐seq Data
http://goo.gl/BW7LFP
"Regions with enriched number of ChIP fragments are potential histone
modification sites or transcription factor binding sites"
My earlier understanding of TFs have them binding upstream of the coding
region and functioning in the process of eventually recruiting polymerase.
In this paper they describe the histone acting in a coding region.
KDM8, a H3K36me2 histone demethylase that acts in the cyclin A1 coding
region to regulate cancer cell proliferation
http://www.pnas.org/content/107/21/9671.full
Part of the problem is with the term "transcription factor". GO doesn't
have a single term for transcription factor.
[Current term] GO:1903506 regulation of nucleic acid-templated transcription
[is_a relation] GO:1903507 negative regulation of nucleic acid-templated
transcription
[is_a relation] GO:1903508 positive regulation of nucleic acid-templated
transcription
[is_a relation] GO:1900259 regulation of RNA-directed RNA polymerase
activity
[is_a relation] GO:0006355 regulation of transcription, DNA-templated
[part_of relation] GO:0000990 transcription factor activity, core RNA
polymerase binding
[part_of relation] GO:0000988 transcription factor activity, protein
binding
Reading the two TF Activity definitions, both say that DNA binding isn't
even a necessary condition for having TF activity. The closest to what I
understand as TF is GO:0000990.
The ENCODE listing at UCSC has a matrix of ChIP-seq with just two
"Transcription factors (TFs) are proteins that bind to DNA and interact
with RNA polymerases to regulate gene expression. Some TFs contain a DNA
binding domain and can bind directly to specific short DNA sequences
('motifs'); others bind to DNA indirectly through interactions with TFs
containing a DNA binding domain. High-throughput antibody capture and
sequencing methods (e.g. chromatin immunoprecipitation followed by
sequencing, or 'ChIP-seq') can be used to identify regions of TF binding
genome-wide. These regions are commonly called ChIP-seq peaks."
If FaceBase needs a further specification of ChIP-seq of chromatin
modifiers then we should have a subtype of "transcription factor binding
site identification by ChIP-Seq assay” rather than a sibling class to avoid
EP300 ChiP-seq being inconsistently assigned to one or the other.
It seems to me that if their intention is to find transcription factor
binding sites it's a different assay than if the intention is to find
histone modifications. I believe different analyses can be used in TF
versus Histone experiments with Chip-Seq. For example, this one: Identifying
differential histone modification sites from ChIP-seq data
http://www.ncbi.nlm.nih.gov/pubmed/22130888.
A last thought: AFAIK it isn't necessarily the case that the site where a
histone binds when acting as a transcription factor (when it has such
activity) is the same as where it binds as a structural element of a
nucleosome.
Maybe the issue should be punted to GO.
Later,
Alan
Related
OBI Terms:
#782Enjoy Italy!!
edit(underlined): It seems to me that if their intention is to find
transcription factor binding sites it's a different assay than if the
intention is to find regions where histone modifiers do their work. I
believe different analyses can be used in TF versus Histone experiments
with Chip-Seq. For example, this one: Identifying differential histone
modification sites from ChIP-seq data
http://www.ncbi.nlm.nih.gov/pubmed/22130888.
I was looking at the OBI hierarchy. Perhaps the definition for chip-Seq is
too tight: an assay which aims at identifying protein binding sites in
genomic DNA and determining how protein may regulate gene transcription
by relying on immunoprecipitation of DNA bound protein, creation of a
library of corresponding DNA fragments (either single or paired-end
fragments) and subsequent sequencing using parallelized sequencing methods.
The underlined: and determining how protein may regulate gene
transcription is the part I think might be worth dropping(or making a
subclass). It seems there are a number of goals that ChIP-seq the method is
uses, not all of which are to directly study gene regulation (i.e. the
specified output is not information about gene expression). A study of
epigenetics could use ChIP-seq to look at issue of inheritence.
So in this paper Epigenetic inheritance uncoupled from sequence-specific
recruitment https://www.sciencemag.org/content/348/6230/1258699.full.pdf they
used the CHiP-seq as a probe for the existence of the histone modifications
in a target region (presumably using antibodies to the modified histones as
pull down). They use a reporter assay to check for the bases that are
modified by them (silencing it). They want to show that just having the
modified histones present is enough to maintain the DNA modifications in
place. They show that DNA modification can be established by binding of a
'starter sequence', but that the starter sequence binding can be removed
and the DNA modifications can persist. They also show a particular domain
in the modified histones isn't necessary to get the modifications
established. Then they show that the domain on the histones (clr4) is
required for the silencing to be maintained, as investigate other proteins
that contribute to maintenance.
So this assay would be something close to histone modification
identification by ChIP-Seq assay, different in that the goal isn't to
discover where the modified histones are but rather to check whether a
specific set of modified histones are in a specified region.
Returning to the request term, after doing yet more poking it looks to me
that while the assay, on the surface, looks like a standard dna binding
assay for EP300, EP300 is classified as a transcriptional coactivator. So
I'm not sure whether the DNA sequenced in the CHiP-seq experiment was
bound to EP300 or to a transcription factor that EP300 binds to. I think it
is the latter, but other opinions would be welcome.
Given that I'll go out on a limb (again) and say this assay isn't a 'A
ChIP-seq assay to identify binding sites for transcription factors'. OTOH,
what's done isn't specific to the function "chromatin modification". Rather
it's a CHiP-seq assay for transcriptional cofactors - i.e. something that
binds to protein that actually binds DNA. Presumably it was determined
elsewhere that EP300 doesn't bind DNA (being a cofactor). In GO it is
annotated with 'regulation of transcription, DNA-templated
http://amigo.geneontology.org/amigo/term/GO:0006355', but not translation
regulator activity, nucleic acid binding
http://amigo.geneontology.org/amigo/term/GO:0090079
One version of a revised hierarchy:
ChIP-seq assay
- chromatin immunoprecipitation with exonuclease sequencing assay (seems
like this should be a defined class as it could be super to any of the
below)
- histone modification identification by ChIP-Seq assay (we know the
modifications in advance)
-- histone modification presence in region by ChIP-Seq assay (we know the
modifications and the region in advance)
- transcription factor binding site identification by ChIP-Seq assay (we
know the TF in advance)
- transcription cofactor activity region identification (We may not know
the TF at all)
-- Chromatin modifier-associated activity region identification (the
requested assay type, cofactor specialized to Chromatin modifier)
Alan
On Mon, Oct 12, 2015 at 6:10 PM Alan Ruttenberg alanruttenberg@gmail.com
wrote:
Related
OBI Terms:
#782I like the structure Alan proposed. The essential distinction (that made me request a new term) is between a factor that directly binds to the DNA, and one that is associated with that region by binding to a protein (that either itself directly binds to DNA or to another chromatin-associated protein...). I tried to send an email early today but I don't see it in the discussion. I've pasted it in below.
Thanks,
Paul.
We struggled with this in the GO as well, which is reflected in the parent term “transcription factor, protein binding” which is more properly considered to be a transcription cofactor. The most common usage of “transcription factor” refers to a DNA binding factor. EP300 does not bind DNA directly (despite the GO annotations that seem to suggest it does), and is thus more properly considered a cofactor that functions via histone acetylation.
Given that the annotation of transcription co-factors is recognized as an issue in GO, I'm OK with the original request that the "chromatin modifier binding site identification by
ChIP-Seq assay" be placed directly under ChIP-seq assay. We should add an editor's note to capture the essence of this discussion.
An example of a chromatin modifier that is not a transcription factor would be sometime that modifies histones such as a histone deacetylase.
Discussed on Nov. 23 call.
currently have
- transcription factor binding site identification by ChIP-Seq assay
now propose to add as sibling. Will be a child of ChIP-Seq assay.
may add a logical axiom adding an objective of identifying transcription cofactor activity region.
- transcription cofactor activity region identification by ChIP-Seq assay
AI: Chris will add.
Added to OBI