Dataset Information

Artificial Intelligence Prediction Across 12,000 Samples Shows Widespread Increased Gene-Gene Chromatin Interactions in Cancers that Constitute Therapeutic Vulnerabilities

ABSTRACT: Gene-gene chromatin interactions (GGIs) bring distal genes into close spatial proximity to permit strong co-expression, which could potentially contribute to cancer progression. High-throughput methods like Hi-C are impractical for very large cohort analyses, thus we developed AI4Loop, an Artificial Intelligence (AI) Deep Learning -based tool to predict GGIs using RNA-Seq data. Applying AI4Loop to 12,000 patient samples from The Cancer Genome Atlas (TCGA) database across 32 cancer types revealed that GGIs show increased cancer sub-type predictivity compared to RNA-Seq data and demonstrated oncogenic gains of GGIs in almost all cancers examined. To target the therapeutic vulnerability of gain of GGIs in cancers, using low-information RNA expression datasets from the CLUE database, we also constructed a drug-perturbation GGI atlas from 50,000 drug-treated samples to identify and repurposed compounds that disrupt oncogenic GGIs. Notably, we found that the antibiotics eperezolid and radezolid reduced cancer-acquired GGIs, which we confirmed with Hi-C experiment. This work showcases AI-directed research in epigenetics, enhances cancer biology predictivity and can promote wide-range drug repurposing in the future. This GEO submission contains the Hi-C data used to test the accuracy of AI4Loop when applied to compounds leading to loss of gene-gene interactions from the CLUE datasets.

ORGANISM(S): Homo sapiens

PROVIDER: GSE287383 | GEO | 2025/01/21

REPOSITORIES: GEO

ACCESS DATA

Dataset's files

Source:

			Action	DRS
		Other

Items per page:

1 - 1 of 1

Similar Datasets

Project description:Background: We have previously shown that the Gene expression Grade Index (GGI) was able to identify two subtypes of estrogen receptor (ER)-positive tumors that were associated with statistically distinct clinical outcomes in both untreated and tamoxifen-treated patients. Here, we aim to investigate the ability of the GGI to predict relapses in postmenopausal women who were treated with tamoxifen (T) or letrozole (L) within the BIG 1-98 trial. Methods: We generated gene expression profiles (Affymetrix) and computed the GGI for a matched, case-control sample of patients enrolled in the BIG 1-98 trial from the two hospitals where frozen samples were available. All relapses (cases) were identified from patients randomized to receive monotherapy or from the switching treatment arms for whom relapse occurred before the switch. Each case was randomly matched with four controls based upon nodal status and treatment (T or L). The prognostic value of GGI was assessed as a continuous predictor and divided at the median. Predictive accuracy of GGI was estimated using time-dependent area under the curve (AUC) of the ROC curves. Results: Frozen samples were analyzable for 48 patients (10 cases and 38 controls). Seven of the 10 cases had been assigned to receive L. Cases and controls were comparable with respect to menopausal and nodal status, local and chemotherapy, and HER2 positivity. Cases were slightly older than controls and had a larger proportion of large, poorly differentiated ER+/PgR- tumors. The GGI was significantly and linearly related to risk of relapse: each 10-unit increase in GGI resulted in an increase of approximately 11% in the hazard rate (p=0.02). Within the subgroups of patients with node-positive disease or who were treated with L, the hazard of relapse was significantly greater for patients with GGI at or above the median. AUC reached a maximum of 78% at 27 months. Conclusions: This analysis supports the GGI as a good predictor of relapse for ER-positive patients, even among patients who receive L. Validation of these results, in a larger series from BIG 1-98, is planned using the simplified GGI represented by a smaller set of genes and tested by qRT-PCR on paraffin-embedded tissues. Test whether the Gene expression Grade Index (GGI) is a potential predictor of relapse for endocrine-treated breast cancer patients in the BIG 1-98 trial. 55 microarray experiments from primary breast tumors of endocrine-treated patients. No replicate, no reference sample.

Dataset Information

Artificial Intelligence Prediction Across 12,000 Samples Shows Widespread Increased Gene-Gene Chromatin Interactions in Cancers that Constitute Therapeutic Vulnerabilities

Dataset's files

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets