JUMP-Target

The JUMP-Target set of perturbations comprises three lists of perturbations

1. JUMP-Target-Compound – a list of compounds with diverse targets.
2. JUMP-Target-ORF – a list of ORF sequences corresponding to genes that are targets of compounds from JUMP-Target-Compound.
3. JUMP-Target-CRISPR – a list of sgRNA sequences corresponding to genes that are targets of compounds from JUMP-Target-Compound.

Both JUMP-Target-ORF and JUMP-Target-CRISPR have sgRNAs and ORFs corresponding to a set of 160 genes, each of which are targets of compounds in JUMP-Target-Compound. JUMP-Target-ORF has perturbations for an additional set of 15 genes that serve as negative controls (below).

Each of the 3 lists fit on a single 384-well plate; the suggested plate layouts are provided (below).

The target annotations were obtained from https://clue.io/repurposing.

This resource was created through the JUMP-Cell Painting Consortium.

Versioning

• There's a single version for the JUMP-Target-CRISPR and JUMP-Target-CRISPR plates (JUMP-Target-1-CRISPR and JUMP-Target-1-CRISPR respectively)
• There are two versions of JUMP-Target-Compound (JUMP-Target-1-Compound and JUMP-Target-2-Compound). Both have the same set of compounds, but have different Broad sample IDs and different layouts. JUMP-Target-2-Compound is used in the production of the JUMP dataset. We recommend using this layout because it will remove layout as a potential confounder when matching against the JUMP dataset.

---

• JUMP-Target-Compound
  • Files
  • Broad internal notes
• JUMP-Target-CRISPR
  • Files
• JUMP-Target-ORF
  • Files

Table of contents generated with markdown-toc

JUMP-Target-Compound

There are 306 compounds in total, of which 46 are included as controls serving different goals

1. poscon_orf: compounds that correlate strongly with (overexpressed) genes in previous Cell Painting experiments. There are 6 such compounds, and their corresponding genes are included in JUMP-Target-ORF and JUMP-Target-CRISPR
2. poscon_cp: compounds with a strong association with the genes that they target, according to ChemicalProbes.org. There are 26 such compounds, and their corresponding genes are included in JUMP-Target-ORF and JUMP-Target-CRISPR.
3. poscon_diverse: pairs of compounds that are strongly correlated to each other, and weakly correlated to other poscon_diverse pairs, in previous Cell Painting experiments. There are 7 such pairs, so 14 compounds in total.
4. negcon: DMSO is the negative control.

The recommended concentration for these compounds is 5 uM.

There are

• n=2 replicate wells for each of 14 poscon_diverse compounds
• n=64 DMSO wells
• n=1 well for the remaining compounds

The company Specs has assembled the compounds for purchase; for info contact [email protected]

Files

JUMP-Target-1_compound_metadata.tsv JUMP-Target-2_compound_metadata.tsv

Column	Description
InChIKey	International Chemical Identifier
pert_iname	Compound name
pubchem_cid	PubChem ID e.g. 72716071
target	Gene target
pert_type	Perturbation type – trt: treatment, control: one of the 3 controls
control_type	Control type – poscon_orf, poscon_cp, or poscon_diverse
smiles	Simplified molecular-input line-entry system (SMILES) string

JUMP-Target-1_compound_platemap.tsv JUMP-Target-2_compound_platemap.tsv

Column	Description
well_position	Well position
broad_sample	Compound ID in Broad Institute's compound management database
solvent	Solvent (always DMSO)

Broad internal notes

• 306 compounds
• JUMP-Target-1_Compound:
  • Compound Request Number: CR-12417 - placed by Niranj Chandrasekaran - $10,404
  • 5 milliMolar 13.75 uL total volume (5 uL dead volume)
• JUMP-Target-2_Compound:
  • Compound Request Number: CR-12908 - placed by Niranj Chandrasekaran - $???
  • 5 milliMolar 13.75 uL total volume (5 uL dead volume)
  • This plate was created because CDoT was unable to create plates with the same plate map as JUMP-Target-1 for the production experiments. Thus, the compounds in JUMP-Target-2 are identical to the compounds in JUMP-Target-1 but the plate maps are different. The broad sample IDs of some compounds are also different.
    • To map the compounds in JUMP-Target-1 to JUMP-Target-2 use InChIKey for "joining".

JUMP-Target-CRISPR

There are 335 sgRNAs (corresponding to 175 genes) in total, of which 88 sgRNAs are included as controls serving different goals

1. poscon_orf: the corresponding genes targets of the poscon_orf compounds. Total: 3 genes x 2 sgRNAs per gene (except for one gene which has a single sgRNA) = 5 sgRNAs.
2. poscon_cp: the corresponding genes targets of the poscon_cp compounds. Total: 13 genes x 2 sgRNAs per gene (except for one gene which has a single sgRNA) = 25 sgRNAs.
3. poscon_diverse: the corresponding genes targets of the poscon_diverse compounds. Total: 14 genes x 2 sgRNAs per gene = 28 sgRNAs.
4. negcon: 30 unique sgRNAs that serve as negative controls because they are either non-targeting (NO_SITE) or target intergenic region (ONE_INTERGENIC_SITE). negcon sgRNAs have an additional annotation e.g. NO_SITE (5 zeros). The number of zeros indicates how deep into the "threat matrix" a particular sgRNA makes it before a single match is found. An sgRNA negcon with higher number of zeros is less likely to have off-target effects and therefore may be a better negative control.

There are

• n=2 replicate wells for each of negcons, as well as for 1 poscon_cp sgRNA, 1 poscon_orf sgRNA, and 13 other sgRNAs
• n=1 well for each remaining sgRNAs
• n=4 empty wells

Files

JUMP-Target-1_crispr_metadata.tsv

Column	Description
broad_sample	sgRNA id in Broad Institute's Genetic Perturbation Platform database
gene	Gene targeted by the sgRNA
pert_type	Perturbation type – trt: treatment, control: one of the 4 controls
control_type	Control type – negcon, poscon_orf, poscon_cp, or poscon_diverse
target_sequence	sgRNA sequence
negcon_control_type	Negative control type – NO_SITE or ONE_INTERGENIC_SITE

JUMP-Target-1_crispr_platemap.tsv

Column	Description
well_position	Well position
broad_sample	sgRNA id in Broad Institute's Genetic Perturbation Platform database

JUMP-Target-ORF

There are 175 ORFs (corresponding to 175 genes) in total, of which 45 ORFs (corresponding to 45 genes) are included as controls serving different goals and 130 genes are represented in duplicate on the plate.

All genes have a single corresponding ORF

1. poscon_orf: the corresponding genes targets of the poscon_orf compounds. Total: 3 genes.
2. poscon_cp: the corresponding genes targets of the poscon_cp compounds. Total: 13 genes.
3. poscon_diverse: the corresponding genes targets of the poscon_diverse compounds. Total: 14 genes.
4. negcon: genes that have had weak signatures in previous Cell Painting overexpression experiments. Total: 15 genes.

There are

• n=2 replicate wells for each of the 30 poscons
• n=4 replicate wells for each of the 15 negcons
• n=2 wells for each remaining ORF (130 of them)
• n=4 empty wells
• For a total of 384 wells

Files

JUMP-Target-1_orf_metadata.tsv

Column	Description
broad_sample	ORF id in Broad Institute's Genetic Perturbation Platform database
gene	Gene overexpressed by the ORF
pert_type	Perturbation type – trt: treatment, control: one of the 4 controls
control_type	Control type – negcon, poscon_orf, poscon_cp, or poscon_diverse

JUMP-Target-1_orf_platemap.tsv

Column	Description
well_position	Well position
broad_sample	ORF id in Broad Institute's Genetic Perturbation Platform database

Compound and gene selection

The list of compounds were derived from Broad's Drug Repurposing Hub dataset, a curated and annotated collection of FDA-approved drugs, clinical trial drugs, and pre-clinical tool compounds. The genes perturbed by genetic perturbations were chosen because they are the annotated targets of the compounds. The Repurposing Hub compounds were filtered using the following criteria:

1. The compounds should target genes that belong to diverse gene families. This is because the ideal methods would work well for many different biological pathways, not just a few that are well-characterized and/or easy to predict.
2. Each gene should be targeted by at least two compounds so that gene-compound matching and compound-compound matching can both be performed using the dataset.
3. The constraint that each compound should target only a single gene, was considered. However, this criterion is difficult to achieve due to polypharmacology, which is the property for compounds to bind and impact many different gene products in the cell; this is especially common for protein kinase inhibitors in the dataset. Instead, only the so-called “historical compounds” listed in the Chemical Probes Portal, comprising compounds that are known to be quite non-selective (or not sufficiently potent) compared with other available chemical probes, were filtered out.
4. Three types of positive control compounds and the genes that they target were selected. (Compound positive controls, CRISPR positive controls, and ORF positive controls).
5. To ensure that the compounds and genes selected were available for performing experiments, compounds that were unavailable for purchase from at least one of four compound vendors (Sigma, SelleckChem, Tocris, and MedChemEx) and genes for which genetic reagents were unavailable from Broad's GPP portal were filtered out.
6. Compounds that belong to DEA's list of controlled substances or OPCW's list of chemical weapons precursors were also excluded.

---

Positive control compounds

Based on our experiments with the Target-1 compounds plates, we have identified a list of eight compounds (and also a subset of four compounds) with maximally diverse phenotypes.

pert_iname	InChIKey	Vendor URL	clue.io URL
AMG900	IVUGFMLRJOCGAS-UHFFFAOYSA-N	https://www.selleckchem.com/products/amg-900.html	https://clue.io/repurposing-app?q=Name:AMG900
NVS-PAK1-1	OINGHOPGNMYCAB-INIZCTEOSA-N	https://www.medchemexpress.com/NVS-PAK1-1.html	https://clue.io/repurposing-app?q=Name:NVS-PAK1-1
dexamethasone	UREBDLICKHMUKA-CXSFZGCWSA-N	https://www.tocris.com/products/dexamethasone_1126	https://clue.io/repurposing-app?q=Name:dexamethasone
LY2109761	IHLVSLOZUHKNMQ-UHFFFAOYSA-N	https://www.selleckchem.com/products/ly2109761.html	https://clue.io/repurposing-app?q=Name:LY2109761
FK-866	KPBNHDGDUADAGP-VAWYXSNFSA-N	https://www.selleckchem.com/products/apo866-fk866.html	https://clue.io/repurposing-app?q=Name:FK-866
quinidine	LOUPRKONTZGTKE-LHHVKLHASA-N	https://www.medchemexpress.com/Quinidine.html	https://clue.io/repurposing-app?q=Name:quinidine
TC-S-7004	CQKBSRPVZZLCJE-UHFFFAOYSA-N	https://www.tocris.com/products/tc-s-7004_5088	https://clue.io/repurposing-app?q=Name:TC-S-7004
aloxistatin	SRVFFFJZQVENJC-IHRRRGAJSA-N	https://www.medchemexpress.com/Aloxistatin.html	https://clue.io/repurposing-app?q=Name:aloxistatin

Compound names in bold is the subset of four compounds.

Layout

Our recommendation is to have at least four replicates of the eight (or four) compounds spread across the plate. If such a layout is not possible, then we recommend the following for a 384-well plate.

pert_iname	well_position
AMG900	B1, J1, F24, N24
NVS-PAK1-1	F1, N1, B24, J24
dexamethasone	C1, K1, G24, O24
LY2109761	E1, M1, A24, I24
FK-866	D1, L1, H24, P24
quinidine	G1, O1, C24, K24
TC-S-7004	H1, P1, D24, L24
aloxistatin	A1, I1, E24, M24

For a 1536-well plate, the layout is similar, but in four quadrants.