Hypoxia classifier for transcriptome datasets

Abstract

Molecular gene signatures are useful tools to characterize the physiological state of cell populations, but most have developed under a narrow range of conditions and cell types and are often restricted to a set of gene identities. Focusing on the transcriptional response to hypoxia, we aimed to generate widely applicable classifiers sourced from the results of a meta-analysis of 69 differential expression datasets which included 425 individual RNA-seq experiments from 33 different human cell types exposed to different degrees of hypoxia (0.1–5%$$\hbox {O}_{2}$$ O 2 ) for 2–48 h. The resulting decision trees include both gene identities and quantitative boundaries, allowing for easy classification of individual samples without control or normoxic reference. Each tree is composed of 3–5 genes mostly drawn from a small set of just 8 genes (EGLN1, MIR210HG, NDRG1, ANKRD37, TCAF2, PFKFB3, BHLHE40, and MAFF). In spite of their simplicity, these classifiers achieve over 95% accuracy in cross validation and over 80% accuracy when applied to additional challenging datasets. Our results indicate that the classifiers are able to identify hypoxic tumor samples from bulk RNAseq and hypoxic regions within tumor from spatially resolved transcriptomics datasets. Moreover, application of the classifiers to histological sections from normal tissues suggest the presence of a hypoxic gene expression pattern in the kidney cortex not observed in other normoxic organs. Finally, tree classifiers described herein outperform traditional hypoxic gene signatures when compared against a wide range of datasets. This work describes a set of hypoxic gene signatures, structured as simple decision tress, that identify hypoxic samples and regions with high accuracy and can be applied to a broad variety of gene expression datasets and formats.

Bhlhe40 limits proliferation and prevents excessive angiogenesis in mouse embryoid bodies under hypoxia

“…We also observed that the knockout of Bhlhe40 or Arnt affected EdU incorporation in normoxic EBs. Applying a hypoxia classifier based on transcriptional changes previously developed by our group to scRNA-seq experiments in EBs, we found that in EBs under hypoxic conditions, the majority of the cells were classified as hypoxic and that in EBs under normoxic conditions, a small number of cells were also classified as hypoxic [45]. In normoxic conditions, the fraction of cells classified as hypoxic may correspond to those located in the core regions of EBs, where oxygen is compromised due to three-dimensional growth.…”

An exploration of vascular cell development in mouse embryoid bodies under hypoxia through single cell RNA-seq

“…• Sheet 1: Cell metadata, including sample ID ("orig.ident"), treatment, and growth time, number of reads and genes detected, percentages of reads assigned to mitochondrial, ribosomal, and chromosome Y, cell cycle scoring results ("S.Score integrated", "G2M.Score integrated", "Phase integrated"), clustering assignment and CUS at a resolution of 0.2 ("res 0.2 Clustering", "res 0.2 Uncertainty"), hypoxia consensus predictions using the tree ensemble described in [35] ("Hx consensusPrediction"), and UMAP reduction coordinates for each cell. • Sheet 2: Clustering assignments for resolutions between 0-1.5.…”

“…Thus, we switched our focus to transcriptional changes occurring due to hypoxia. First, to evaluate the global effects of the treatment on the EBs we run the hypoxia classifier previously developed [35], using as input the raw counts of genes detected in the dataset. The results are displayed on a UMAP projection of the dataset in Fig.…”

“…6 Application of the hypoxia classifier. Hypoxia scores derived from the application of a tree ensemble classifier predicting whether cells were exposed to hypoxia [35]. Consensus scores were calculated as the mean probability of each cell's transcriptome to be hypoxic given by each of the 20 trees in the ensemble.…”