The central challenge the researchers set out to address is one that has frustrated computational biologists for years: systematically identifying transcriptional and epigenetic regulators, or TERs, in myeloid leukemia. Most existing approaches rely on a single type of data, such as gene expression alone, and they struggle with one of the defining features of human genomic architecture, namely long-range regulation. Enhancers, the DNA elements that boost gene activity, can sit hundreds of thousands of base pairs away from the promoters they control, often separated by vast stretches of non-coding sequence. Without a reliable way to link enhancers to their target promoters, regulatory analysis remains incomplete, and candidate disease drivers slip through the net.

TERfinder takes a fundamentally different approach by integrating multi-omics features within a deep learning architecture. Rather than asking what a single data type says about a genomic region, the framework fuses information across layers of cellular regulation, including histone modification signals that mark active or primed regulatory elements. The model’s task is to predict enhancer-promoter interactions, or EPIs, the physical and functional connections through which regulatory elements control gene expression. By learning the patterns that distinguish true interactions from chance proximity, the model builds a map of the regulatory circuitry operating in leukemia cells.

The performance figures reported in the study are striking. On held-out chromosomes, meaning genomic regions the model had never seen during training, TERfinder achieved an area under the receiver operating characteristic curve of 0.9644 and an area under the precision-recall curve of 0.9584. These numbers indicate that the framework can discriminate interacting enhancer-promoter pairs from non-interacting ones with remarkable accuracy even on unfamiliar territory. Crucially, the authors compared their model against baseline versions stripped of either the autoencoder component or the histone features, and statistical testing using the DeLong method confirmed that the full model significantly outperformed these ablated versions, with p-values below 0.01. That comparison matters because it demonstrates that the multi-omics integration is not decorative; each layer of information contributes measurably to predictive power.

With the interaction predictions in hand, the researchers turned to the question of which regulators matter most in myeloid leukemia. Motif enrichment analysis, which searches for the DNA-binding signatures of transcription factors within predicted regulatory regions, flagged members of the C/EBP and ETV families as candidate master regulators. These findings were not left as abstract computational output. The team performed single-cell regulon analysis, examining regulatory network activity at the resolution of individual cells, and confirmed that these transcription factor families are indeed active in acute myeloid leukemia progenitor populations, the immature cells thought to seed and sustain the disease.

The single-cell analysis yielded one of the study’s most compelling results: regulatory networks centered on the transcription factors SPI1 and CEBPA were found to be active in AML blasts, the malignant cells that crowd out healthy blood production in patients. SPI1, also known as PU.1, is a well-established lineage-determining factor in myeloid cells, and CEBPA is a transcription factor whose mutations are a recognized driver of a distinct AML subtype. The finding that their regulon activity, the collective output of the gene networks they control, is associated with poor overall survival elevates these networks from biological curiosities to clinically meaningful signatures of aggressive disease.

Perhaps the most translationally significant result is a compact four-gene expression signature comprising SPI1, CEBPA, MYC, and PTPN6. When the researchers stratified AML patients according to the activity of this signature, they separated them into high-risk and low-risk groups with a log-rank p-value below 0.01, a standard statistical threshold in survival analysis. A four-gene panel is the kind of lean biomarker that could plausibly be adapted into clinical assays, offering a way to refine risk stratification beyond the cytogenetic and mutational markers currently in use. MYC, the famed oncogenic transcription factor, and PTPN6, a phosphatase gene involved in immune cell signaling, round out a signature that spans both classical cancer biology and myeloid lineage regulation.

The methodological significance of TERfinder extends beyond leukemia. The framework demonstrates a general strategy for regulatory inference in which deep learning serves as the integrator of heterogeneous genomic signals, converting the raw complexity of multi-omics data into interpretable predictions about which regulatory elements talk to which genes. The authors describe it as a framework for multi-omics regulatory inference and candidate transcription factor identification in myeloid leukemia, and the architecture’s reliance on features that are broadly measurable, such as histone marks and chromatin accessibility proxies, suggests the approach could be retrained for other cancer types or developmental contexts where enhancer-promoter wiring goes awry.

The study also reflects a broader shift in how computational genomics is approaching cancer. For decades, the field has catalogued mutations and expression changes, but the regulatory layer, the system that decides when and how strongly genes are turned on, has been harder to read. Tools like TERfinder represent an attempt to make that layer computationally legible. By predicting enhancer-promoter interactions with high accuracy, identifying the transcription factors that occupy the resulting regulatory hubs, and connecting those hubs to patient survival, the framework closes a loop from DNA sequence to clinical outcome that few methods have managed to traverse.

There remain, of course, the usual caveats that accompany any computational framework. The model’s predictions are inferences derived from training data, and experimental validation of individual enhancer-promoter links will be needed before the full interaction map can be treated as ground truth. The four-gene signature, while statistically robust in the analyzed cohorts, would require independent validation before clinical deployment. Yet the open-access publication, the extensive supplementary tables released alongside the paper, and the framework’s demonstrated superiority over simplified baselines give the work a solid foundation. For a disease as relentless as acute myeloid leukemia, where five-year survival remains unacceptably low for many patient groups, a tool that can spotlight the regulatory circuits sustaining the cancer, and distill them into a measurable signature of risk, is a development worth watching closely.

Subject of Research: Deep learning-based multi-omics analysis of transcriptional and epigenetic regulation in myeloid leukemia

Article Title: TERfinder: A deep learning framework for multi-omics regulatory analysis in myeloid leukemia

Article References: Xu, M., Xu, X., Yufei, L., Zhang, G., Liu, J., Cui, T., Guo, B., Huang, J., & Li, C. (2026). TERfinder: A deep learning framework for multi-omics regulatory analysis in myeloid leukemia. BMC Bioinformatics. https://doi.org/10.1186/s12859-026-06627-5

Image Credits: AI Generated

DOI: 10.1186/s12859-026-06627-5

Keywords: TERfinder, deep learning, multi-omics, enhancer-promoter interactions, acute myeloid leukemia, transcriptional regulation, epigenomics, gene regulatory networks, SPI1, CEBPA, biomarker, single-cell analysis

The research, led by Min Wu, Yi-Fei Yang and Xiang-Ping Du of the Department of Orthopedics at Jiangxi Provincial Children’s Hospital in Nanchang, China, began with a simple observation that belied a complex mechanism. When the team exposed BMSCs to hydrogen peroxide, a standard laboratory approach for mimicking oxidative stress, the levels of SPI1, a transcription factor encoded by the Spi-1 proto-oncogene, dropped sharply. The same decline was visible in an animal model: rats whose ovaries had been removed to induce estrogen-deficiency osteoporosis also showed reduced SPI1 in their bone marrow-derived mesenchymal stem cells. The parallel was striking. In both the culture dish and the diseased skeleton, oxidative stress appeared to strip the cells of a factor they apparently needed to stay young.

To test whether that loss mattered, the researchers forced SPI1 expression back up in stressed cells. The effect was dramatic. Overexpressing SPI1 rescued the senescent phenotype that hydrogen peroxide had induced: the cells regained viability, produced fewer reactive oxygen species, and showed reduced staining for senescence-associated beta-galactosidase, a classic enzymatic marker of cellular aging. Just as importantly, SPI1 restoration preserved the cells’ ability to differentiate into bone-forming osteoblasts, which the team quantified using alizarin red S staining of mineralized matrix. In other words, SPI1 was not simply a bystander that declined as cells aged; it appeared to be an active guardian of both youth and bone-building function in the mesenchymal stem cell compartment.

The next question was how SPI1 exerted this protection. Transcription factors work by binding DNA and switching genes on or off, and the researchers had a specific target in mind: NAMPT, the gene encoding nicotinamide phosphoribosyltransferase, the rate-limiting enzyme of the NAD+ salvage pathway. NAD+ has become one of the most intensively studied molecules in aging biology, serving as a coenzyme for hundreds of redox reactions and as the fuel consumed by sirtuins, a family of deacylating enzymes that regulate stress resistance, mitochondrial function and longevity. Declining NAD+ levels are a hallmark of aged tissues, and boosting NAMPT activity has previously been shown to rejuvenate senescent mesenchymal stem cells. Through chromatin immunoprecipitation and luciferase reporter assays, the team demonstrated that SPI1 binds directly to the NAMPT promoter and transcriptionally activates it, raising NAD+ levels inside the stressed cells. This elevation of NAD+ was the mechanistic bridge between SPI1 and senescence resistance, connecting a transcription factor to one of aging biology’s most celebrated metabolic pathways.

But SPI1 is itself a protein, and proteins in the cellular environment are subject to constant quality control. The ubiquitin-proteasome system tags unwanted proteins with chains of ubiquitin molecules, marking them for destruction. For a transcription factor like SPI1, ubiquitination can mean a short half-life and a rapid collapse of expression under stress. This is where USP22 enters the story. USP22 is a deubiquitinating enzyme, a member of a large family of proteases that remove ubiquitin tags and thereby rescue proteins from degradation. The enzyme has attracted growing attention in recent years for its role in stabilizing key regulatory proteins: it has been shown, for example, to deubiquitinate PPARγ in liver cancer cells and FoxM1 in endometrial stromal cells, and to suppress inflammatory signaling by influencing the NLRP3 inflammasome. The Jiangxi team hypothesized that USP22 might perform the same service for SPI1 in BMSCs.

The experimental evidence supported the hypothesis at every step. Co-immunoprecipitation confirmed that USP22 physically interacts with SPI1 inside the cells. Ubiquitination assays then showed that USP22 removes ubiquitin chains from SPI1, and cycloheximide chase experiments, which block new protein synthesis and reveal how fast existing proteins decay, demonstrated that SPI1 protein persists far longer when USP22 is abundant. With SPI1 stabilized, NAMPT transcription stayed high, NAD+ levels were maintained, and the senescence program was held in check. The final piece of the causal puzzle came from a knockdown experiment: when the researchers silenced SPI1 in cells overexpressing USP22, the protective effects of USP22 largely evaporated. Senescence markers rose again, and osteogenic differentiation faltered, showing that SPI1 is the essential downstream mediator of USP22’s anti-aging action in these cells.

Why does this circuit matter beyond the laboratory? Osteoporosis, which affects hundreds of millions of people worldwide, is fundamentally a disease of failed bone regeneration, in which the balance tips from bone formation to bone resorption. Estrogen deficiency accelerates oxidative stress in the bone marrow, and senescent BMSCs not only fail to produce new osteoblasts but actively worsen the inflammatory milieu. The ovariectomized rat model used in this study reproduces key features of postmenopausal bone loss, and the finding that SPI1 declines in this model suggests that the USP22-SPI1-NAMPT axis may be a plausible therapeutic target. If a drug or gene therapy could raise USP22 activity, or stabilize SPI1 by another means, or simply replenish NAD+ through well-known precursors such as nicotinamide mononucleotide, the senescence cascade in transplanted or endogenous stem cells might be blunted before it begins.

The findings also connect to a much broader scientific conversation about the biology of aging. Cellular senescence was once viewed as a passive consequence of damage accumulation; it is now understood as a tightly regulated program, orchestrated by p53, p21, p16 and chromatin changes, and driven upstream by metabolic deficits such as NAD+ depletion. Sirtuins sit at the junction of metabolism and chromatin, consuming NAD+ as they modify histones and regulatory proteins, which is why the NAD+-sirtuin axis has become a focal point for anti-aging interventions. What this study adds is an upstream control point: a deubiquitinase-transcription factor module that feeds into NAMPT expression and thereby into the entire NAD+ economy of the cell. It suggests that protein stability, not merely gene expression or enzyme activity, may be a decisive lever for keeping stem cells youthful under stress.

There are, of course, important caveats and open questions. USP22 has a complicated double life in the literature: in some contexts it promotes tumor growth and inflammation, and its substrates are numerous and tissue-dependent. Any clinical strategy that elevates USP22 would need to weigh potential oncogenic risks, particularly since SPI1 itself is best known as an oncogene in hematopoietic malignancies, even as the new work assigns it a protective role in mesenchymal stem cells. The current study also relies on cell culture and a rat model; whether the same circuit operates in human bone marrow, and whether it can be safely modulated in patients, remains to be demonstrated. Still, the elegance of the pathway is hard to ignore: oxidative stress degrades SPI1, loss of SPI1 silences NAMPT, silencing of NAMPT drains NAD+, and drained NAD+ leaves stem cells old and bone barren. By inserting USP22 at the top of that cascade, evolution appears to have built a safeguard, and researchers have now found the switch. For an aging global population facing a rising tide of fractures, that switch is worth pulling.

Subject of Research: How the deubiquitinase USP22 protects bone marrow mesenchymal stem cells from oxidative stress-induced senescence

Article Title: USP22 alleviates oxidative stress-induced BMSCs senescence by stabilizing SPI1 protein

Article References: Wu, M., Yang, Y.-F., & Du, X.-P. (2026). USP22 alleviates oxidative stress-induced BMSCs senescence by stabilizing SPI1 protein. Molecular Genetics and Genomics, 301(1), Article 194. https://doi.org/10.1007/s00438-026-02519-8

Image Credits: AI Generated

DOI: 10.1007/s00438-026-02519-8

Keywords: USP22, SPI1, NAMPT, NAD+, bone marrow mesenchymal stem cells, cellular senescence, oxidative stress, deubiquitination, osteoporosis, osteogenic differentiation, sirtuins, protein stability