FOXO3/TGFB short mention (CRISPR)

afermg commented 10 months ago

Moved from https://github.com/jump-cellpainting/morphmap/issues/160

From a mail chain with some startup looking to use FOXO3 as a therapeutic target.

Probably like two sentences in the final paper saying something like "We first looked at well-known proteins, starting with TGFb, and we saw FOXO3 and a bunch of other well-known pathway members and interactors were strongly correlated/anto-correlated (fig x)"

    have compounds that impact FOXO3 but not sure what is the direct target
    Him seeing TGFb on the bottom of the list of FOXO3 is very exciting because involved in fibrosis
    FOXO3/TGFb relationship well known: totally fine to include this vignette in MorphMap paper, not confidential

Found an already validated anticorrelated pair, we should add a bit of info on this FOXO3/TGFB business. @AnneCarpenter AnneCarpenter commented Jan 24, 2024

Is it possible to move this to 2023_12_JUMP_data_only_vignettes repo? I didn't have permissions (maybe it's not possible). That will help me to track it better. @afermg

AnneCarpenter commented 10 months ago

@afermg Are you willing and eager to write up this chunk for the paper (which is here: https://docs.google.com/document/d/160QYCeJXMJMPgvYcirkkKaY99r6dZq6FMqYGE61CfFc/edit - search for ALANHERE for the right spot) Sometime in the next ~3 weeks? If not, I can give it a shot.

afermg commented 10 months ago

Do you mean the one before or after "ALANHERE"? If the former, I am unsure about which Knowledge Graph that is referring to; if the latter, is "Genes likely to have Cell Painting phenotypes share certain characteristics" referring to the feature_exploration query tool I'm working on?

AnneCarpenter commented 10 months ago

Ah, your text would stand alone, it just goes between those two sections of existing text. If I were less rushed i'd have suggested a heading for your new chunk, something like "Morphological similarities reveal expected gene connections"

afermg commented 10 months ago

Noted, I can give it a go

On Wed, Jan 24, 2024 at 1:39 PM Anne Carpenter @.***> wrote:

Ah, your text would stand alone, it just goes between those two sections of existing text. If I were less rushed i'd have suggested a heading for your new chunk, something like "Morphological similarities reveal expected gene connections"

— Reply to this email directly, view it on GitHub https://github.com/broadinstitute/2023_12_JUMP_data_only_vignettes/issues/22#issuecomment-1908711441, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADNQL2HFCEY4YTO6AN6VCCLYQFIMLAVCNFSM6AAAAABCJEOU3KVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMBYG4YTCNBUGE . You are receiving this because you were mentioned.Message ID: <broadinstitute/2023_12_JUMP_data_only_vignettes/issues/22/1908711441@ github.com>

AnneCarpenter commented 10 months ago

Do you recall if this finding was in ORF or CRISPR? My guess is it's whichever one you made the website for first. I'm adding (ORF) or (CRISPR) to the end of each vignette issue title so we can keep track of which ones need reconsidering in light of CRISPR profiles changing.

afermg commented 10 months ago

It was in CRISPR. I think there are tags in github that may be useful for that purpose.

AnneCarpenter commented 10 months ago

ok, thanks - and good point! but too late now :D

niranjchandrasekaran commented 6 months ago

In the latest ORF and CRISPR profiles, the FOXO3-TGFB connection is no longer present in both ORF and CRISPR.

Will reopen this issue if we are still interested in this relationship.

AnneCarpenter commented 6 months ago

Have you updated the manuscript accordingly? I'd feel better knowing if this dropped out because one or the other genes fell below 'has a phenotype' threshold, or if somehow the correlation suddenly dropped too low :(

niranjchandrasekaran commented 6 months ago

connections

Update: The gene's name turns out to be TGFB1 and not TGFB. Here are the results below

ORF

FOXO3 does not have a phenotype in ORF while TGFB1 does.

CRISPR

Both genes have a phenotype, but their cosine similarity is only -0.21

Though the similarity is weak, there is some signal so I will reopen this issue.

niranjchandrasekaran commented 6 months ago

Notebook

The heatmap shows the percentile of the cosine similarities (1 → similar, 0 → anti-similar). The text is the maximum of the absolute KG score (gene_mf__go, gene_bp_go, gene_pathway). I set a KG threshold (like we previously had) of 0.4. If connections have a score lesser than this threshold, then the connection is considered to be unknown. The KG scores were downloaded from Google Drive: ORF and CRISPR. The diagonal of the heatmap indicates whether a gene has a phenotype (False could also mean the gene is not present in the dataset).