Heterogeneity Of Breast Cancer Research Articles

Multi-omics analyses unravel DNA damage repair-related clusters in breast cancer with experimental validation.

As one of the most common malignancies worldwide, breast cancer (BC) exhibits high heterogeneity of molecular phenotypes. The evolving view regarding DNA damage repair (DDR) is that it is context-specific and heterogeneous, but its role in BC remains unclear. Multi-dimensional data of transcriptomics, genomics, and single-cell transcriptome profiling were obtained to characterize the DDR-related features of BC. We collected 276 DDR-related genes based on the Molecular Signature Database (MSigDB) database and previous studies. We acquired public datasets included the SCAN-B dataset (GEO: GSE96058), METABRIC database, and TCGA-BRCA database. Corresponding repositories such as transcriptomics, genomics, and clinical information were also downloaded. We selected scRNA-seq data from GEO: GSE176078, GSE114727, GSE161529, and GSE158724. Bulk RNA-seq data from GEO: GSE176078, GSE18728, GSE5462, GSE20181, and GSE130788 were extracted for independent analyses. The DDR classification was constructed in the SCAN-B dataset (GEO: GSE96058) and METABRIC database, Among BC patients, there were two clusters with distinct clinical and molecular characteristics: the DDR-suppressed cluster and the DDR-active cluster. A superior survival rate is found for tumors in the DDR-suppressed cluster, while those with the DDR-activated cluster tend to have inferior prognoses and clinically aggressive behavior. The DDR classification was validated in the TCGA-BRCA cohort and shown similar results. We also found that two clusters have different pathway activities at the genomic level. Based on the intersection of the different expressed genes among these cohorts, we found that PRAME might play a vital role in DDR. The DDR classification was then enabled by establishing a DDR score, which was verified through multilayer cohort analysis. Furthermore, our results revealed that malignant cells contributed more to the DDR score at the single-cell level than nonmalignant cells. Particularly, immune cells with immunosuppressive properties (such as FOXP3+ CD4+ T cells) displayed higher DDR scores among those with distinguishable characteristics. Collectively, this study performs general analyses of DDR heterogeneity in BC and provides insight into the understanding of individualized molecular and clinicopathological mechanisms underlying unique DDR profiles.

Organ-Specificity of Breast Cancer Metastasis.

Breast cancer (BC) remains one of the most common malignancies among women worldwide. Breast cancer shows metastatic heterogeneity with priority to different organs, which leads to differences in prognosis and response to therapy among patients. The main targets for metastasis in BC are the bone, lung, liver and brain. The molecular mechanism of BC organ-specificity is still under investigation. In recent years, the appearance of new genomic approaches has led to unprecedented changes in the understanding of breast cancer metastasis organ-specificity and has provided a new platform for the development of more effective therapeutic agents. This review summarises recent data on molecular organ-specific markers of metastasis as the basis of a possible therapeutic approach in order to improve the diagnosis and prognosis of patients with metastatically heterogeneous breast cancer.

‘Inner world’ of luminal HER 2-negative metastatic breast cancer: Efficacy of ribociclib in different intrinsic subtypes

The heterogeneity of breast cancer (BC), which determines various clinical scenarios of the disease, is still inaccessible to our understanding. Heterogeneity is based primarily on intrinsic subtypes of breast cancer, determined using genomic tests. Currently, four intrinsic subtypes are generally recognized: luminal A and B, HER-enriched (HER2-E) and basal-like. HER2-E subtype in luminal HER-negative (ER+HER–) breast cancer accounts for 11–22 per cent of cases. The efficacy of a combination of hormone therapy (HT) and the CDK4/6 inhibitor ribociclib in the HER2-E subtype was evaluated in a large exploratory analysis of MONALEESA-2, -3, and -7 trials. The gain in progression-free survival (PFS) and overall survival (OS) from the addition of ribociclib to HT was observed in all intrinsic subtypes, except for basal-like, and reached maximum in HER2-E subtype: Ribociclib increased the median PFS to 16.4 months from 5.5 months on HT (HR = 0.389; p < 0.0001) and median OS to 40.3 months from 29.4 months (HR = 0.600; p = 0.0180). The significant frequency of occurrence of the HER2-E subtype in ER+HER-BC, the low efficiency of monoHT in this subtype is another argument in favor of prescribing combination therapy in early lines. Information about the intrinsic subtype not only brings us closer to understanding tumor heterogeneity, but may also become a significant factor in determining treatment tactics in the future.

Multi-omics integration analysis unveils heterogeneity in breast cancer at the individual level

ABSTRACT Identifying robust breast cancer subtypes will help to reveal the cancer heterogeneity. However, previous breast cancer subtypes were based on population-level quantitative gene expression, which is affected by batch effects and cannot be applied to individuals. We detected differential gene expression, genomic, and epigenomic alterations to identify driver differential expression at the individual level. The individual driver differential expression reflected the breast cancer patients’ heterogeneity and revealed four subtypes. Mesenchymal subtype as the most aggressive subtype harbored deletion and downregulated expression of genes in chromosome 11q23 region. Specifically, silencing of the SDHD gene in 11q23 promoted the invasion and migration of breast cancer cells in vitro by the epithelial–mesenchymal transition. The immunologically hot subtype displayed an immune-hot microenvironment, including high T-cell infiltration and upregulated PD-1 and CTLA4. Luminal and genomic-unstable subtypes showed opposite macrophage polarization, which may be regulated by the ligand–receptor pairs of CD99. The integration of multi-omics data at the individual level provides a powerful framework for elucidating the heterogeneity of breast cancer.

KLHL29-mediated DDX3X degradation promotes chemosensitivity by abrogating cell cycle checkpoint in triple-negative breast cancer

Triple-negative breast cancer (TNBC) is a heterogeneous breast cancer subtype and accounts for approximately 15–20% of breast cancer cases. In this study, we identified KLHL29, which is an understudied member of the Kelch-like gene family, as a crucial tumor suppressor that regulates chemosensitivity in TNBC. KLHL29 expression was significantly downregulated in breast cancer tissues compared with adjacent normal tissues, and low levels of KLHL29 were associated with unfavorable prognoses. Ectopic KLHL29 suppressed, while depleting KLHL29 promoted, the growth, proliferation, migration, and invasion of TNBC. Mechanistically, KLHL29 recruited the CUL3 E3-ligase to the RNA-binding protein DDX3X, leading to the proteasomal degradation of the latter. This downregulation of DDX3X resulted in the destabilization of CCND1 mRNA and the consequent cell cycle arrest at G0/G1 phase. Remarkably, the DDX3X inhibitor RK33 combined with platinum-based chemotherapy can synergistically suppress TNBC that usually expresses low levels of KLHL29 and high levels of DDX3X using cancer cell-derived xenograft and patient-derived organoids models. Altogether, we uncovered the potential role for the KLHL29-DDX3X signaling cascade in the regulation of TNBC progression, thus providing a promising combination strategy for overcoming TNBC chemoresistance.

BGSSN: Breast Cancer-Associated Genes Prediction Based on Weighted Sample-Specific Networks of Cancer Subtypes

Breast cancer exhibits a notable degree of heterogeneity in its occurrence and progression, encompassing diverse clinical patterns and outcomes among patients even with identical clinical pathological stages. Genetic mutations in different subtypes of breast cancer may lead to different types of disease and have different clinical implications. Therefore, molecular typing based on the characteristics of breast cancer heterogeneity and the screening of associated genes for different subtypes of breast cancer may be able to more accurately determine the pathogenic genes of breast cancer. In this paper, we propose a weighted sample-specific network based on breast cancer subtypes to predict associated genes, named BGSSN. To better reflect the individual characteristics of patients and the importance of patient samples in different subtypes, the weight of samples is added when constructing the sample-specific network. The random walk with restart method is then utilized to predict new breast cancer-associated genes within the constructed network. By leveraging this method, the network structure can be effectively explored to identify potential gene candidates.

Integrative multi-omic sequencing reveals the MMTV-Myc mouse model mimics human breast cancer heterogeneity

BackgroundBreast cancer is a complex and heterogeneous disease with distinct subtypes and molecular profiles corresponding to different clinical outcomes. Mouse models of breast cancer are widely used, but their relevance in capturing the heterogeneity of human disease is unclear. Previous studies have shown the heterogeneity at the gene expression level for the MMTV-Myc model, but have only speculated on the underlying genetics.MethodsTumors from the microacinar, squamous, and EMT histological subtypes of the MMTV-Myc mouse model of breast cancer underwent whole genome sequencing. The genomic data obtained were then integrated with previously obtained matched sample gene expression data and extended to additional samples of each histological subtype, totaling 42 gene expression samples. High correlation was observed between genetic copy number events and resulting gene expression by both Spearman’s rank correlation coefficient and the Kendall rank correlation coefficient. These same genetic events are conserved in humans and are indicative of poor overall survival by Kaplan–Meier analysis. A supervised machine learning algorithm trained on METABRIC gene expression data was used to predict the analogous human breast cancer intrinsic subtype from mouse gene expression data.ResultsHerein, we examine three common histological subtypes of the MMTV-Myc model through whole genome sequencing and have integrated these results with gene expression data. Significantly, key genomic alterations driving cell signaling pathways were well conserved within histological subtypes. Genomic changes included frequent, co-occurring mutations in KIT and RARA in the microacinar histological subtype as well as SCRIB mutations in the EMT subtype. EMT tumors additionally displayed strong KRAS activation signatures downstream of genetic activating events primarily ascribed to KRAS activating mutations, but also FGFR2 amplification. Analogous genetic events in human breast cancer showed stark decreases in overall survival. In further analyzing transcriptional heterogeneity of the MMTV-Myc model, we report a supervised machine learning model that classifies MMTV-Myc histological subtypes and other mouse models as being representative of different human intrinsic breast cancer subtypes.ConclusionsWe conclude the well-established MMTV-Myc mouse model presents further opportunities for investigation of human breast cancer heterogeneity.

OR33-01 Live Cell Imaging Of Enhancer Transcriptional States

Abstract Disclosure: F. Chapus: None. C. Day: None. P. Yasar: None. C. Bregio: None. A. Palia: None. E. Scappini: None. C. Tucker: None. J. Rodriguez: None. Intra-tumoral gene expression heterogeneity of estrogen receptor positive breast cancers is a roadblock toward effective treatment due to subpopulations acquiring resistance and metastatic ability. This expression heterogeneity can be explained by the stochastic nature of transcription. Transcriptional initiation is coordinated through distal elements known as enhancers. These elements aid in the recruitment of cofactors and chromatin remodelers to promoters. During transcriptional activation, genes are transcribed during short periods of nascent RNA synthesis known as transcriptional bursts. Furthermore, estrogen responsive genes can transition between transcriptional bursting and variably long periods of inactivity lasting from minutes to days. Here, we investigated whether estrogen responsive enhancers of variably expressed genes also exhibited transcriptional states which reflect promoter states. To observe live cell transcription dynamics, we integrated MS2 stem loops into several endogenous enhancers using CRISPR. After transcription initiation, these stem loops are transcribed and subsequently bound by the RNA binding protein, MS2-GFP. This enrichment in MS2-GFP at the transcription site is visible as a bright punctate spot. We observed that the estrogen responsive TFF1 enhancer exhibited short periods of transcriptional activity and long periods of inactivity. Moreover, the TFF1 enhancer RNA (eRNA) sporadically formed dynamic, but rare and long-lived condensates. To correlate enhancer and promoter transcription we used single molecule fluorescence in situ hybridization of several eRNA. We observed that eRNA aggregates were correlated with an increase of mRNA content and occasionally localized to different nuclear bodies. Lastly, we enriched for chromatin bound RNA and observed long enhancer RNA transcripts. Taken together, our work highlights the dynamic coupling of enhancer and promoter regulation and the importance of understanding the regulators of enhancer transcriptional states in cancer. Presentation: Sunday, June 18, 2023

NKG2A Is a Therapeutic Vulnerability in Immunotherapy Resistant MHC-I Heterogeneous Triple-Negative Breast Cancer.

Clinical resistance to immunotherapy is common in breast cancer, and many patients will likely require combination therapy to maximize immunotherapeutic benefit. This study demonstrates that heterogeneous MHC-I expression drives resistance to anti-PD-L1 therapy and exposes NKG2A on NK cells as a target to overcome resistance. This article is featured in Selected Articles from This Issue, p. 201.

Identifying the effectiveness of 3D culture systems to recapitulate breast tumor tissue in situ.

Breast cancer heterogeneity contributes to chemotherapy resistance and decreased patient survival. To improve patient outcomes it is essential to develop a technology that is able to rapidly select the most efficacious therapy that targets the diverse phenotypes present within the tumor. Breast cancer organoid technologies are proposed as an attractive approach for evaluating drug responses prior to patient therapy. However, there remain challenges in evaluating the effectiveness of organoid cultures to recapitulate the heterogeneity present in the patient tumor in situ. Organoids were generated from seven normal breast and nineteen breast cancer tissues diagnosed as estrogen receptor positive or triple negative. The Jensen-Shannon divergence index, a measure of the similarity between distributions, was used to compare and evaluate heterogeneity in starting tissue and their resultant organoids. Heterogeneity was analyzed using cytokeratin 8 and cytokeratin 14, which provided an easily scored readout. In the in vitro culture system HER1 and FGFR were able to drive intra-tumor heterogeneity to generate divergent phenotypes that have different sensitivities to chemotherapies. Our methodology, which focuses on quantifiable cellular phenotypes, provides a tractable system that complements omics approaches to provide an unprecedented view of heterogeneity and will enhance the identification of novel therapies and facilitate personalized medicine.

Phenotypic heterogeneity drives differential disease outcome in a mouse model of triple negative breast cancer.

The triple negative breast cancer (TNBC) subtype is one of the most aggressive forms of breast cancer that has poor clinical outcome and is an unmet clinical challenge. Accumulating evidence suggests that intratumoral heterogeneity or the presence of phenotypically distinct cell populations within a tumor play a crucial role in chemoresistance, tumor progression and metastasis. An increased understanding of the molecular regulators of intratumoral heterogeneity is crucial to the development of effective therapeutic strategies in TNBC. To this end, we used an unbiased approach to identify a molecular mediator of intratumoral heterogeneity in breast cancer by isolating two tumor cell populations (T1 and T2) from the 4T1 TNBC model. Phenotypic characterization revealed that the cells are different in terms of their morphology, proliferation and self-renewal ability in vitro as well as primary tumor formation and metastatic potential in vivo. Bioinformatic analysis followed by Kaplan Meier survival analysis in TNBC patients identified Metastasis associated colon cancer 1 (Macc1) as one of the top candidate genes mediating the aggressive phenotype in the T1 tumor cells. The role of Macc1 in regulating the proliferative phenotype was validated and taken forward in a therapeutic context with Lovastatin, a small molecule transcriptional inhibitor of Macc1 to target the T1 cell population. This study increases our understanding of the molecular underpinnings of intratumoral heterogeneity in breast cancer that is critical to improve the treatment of women currently living with the highly aggressive TNBC subtype.

Metabolic stratification of human breast tumors reveal subtypes of clinical and therapeutic relevance

Extensive metabolic heterogeneity in breast cancers has limited the deployment of metabolic therapies. To enable patient stratification, we studied the metabolic landscape in breast cancers (∼3000 patients combined) and identified three subtypes with increasing degrees of metabolic deregulation. Subtype M1 was found to be dependent on bile-acid biosynthesis, whereas M2 showed reliance on methionine pathway, and M3 engaged fatty-acid, nucleotide, and glucose metabolism. The extent of metabolic alterations correlated strongly with tumor aggressiveness and patient outcome. This pattern was reproducible in independent datasets and using invivo tumor metabolite data. Using machine-learning, we identified robust and generalizable signatures of metabolic subtypes in tumors and cell lines. Experimental inhibition of metabolic pathways in cell lines representing metabolic subtypes revealed subtype-specific sensitivity, therapeutically relevant drugs, and promising combination therapies. Taken together, metabolic stratification of breast cancers can thus aid in predicting patient outcome and designing precision therapies.

A Switch-On Affinity-Based Iridium(III) Conjugate Probe for Imaging Mitochondrial Glutathione S-Transferase in Breast Cancer Cells.

Glutathione S-transferase is heterogeneously expressed in breast cancer cells and is therefore emerging as a potential diagnostic biomarker for studying the heterogeneity of breast cancers. However, available fluorescent probes for GSTs depend heavily on GSTs-catalyzed glutathione (GSH) nucleophilic substitution reactions, making them susceptible to interference by the high concentration of nucleophilic species in the cellular environment. Moreover, the functions of subcellular GSTs are generally overlooked due to the lack of suitable luminescence probes. Herein, we report a highly selective affinity-based luminescence probe 1 for GST in breast cancer cells through tethering a GST inhibitor, ethacrynic acid, to an iridium(III) complex. Compared to activity-based probes which require the use of GSH, this probe could image GST-pi in the mitochondria by directly adducting to GST-pi (or potentially GST-pi/GS) in living cells. Probe 1 possesses desirable photophysical properties including a lifetime of 911 ns, a Stokes shift of 343 nm, and high photostability. The "turn on" luminescence mode of the probe enables highly selective detection of the GST with a limit of detection of 1.01 μM, while its long emission lifetime allows sensitive detection in organic dye-spiked autofluorescence samples by a time-resolved mode. The probe was further applied to specifically and quantitatively visualize MDA-MB-231 cells via specific binding to mitochondrial GST, and could differentiate breast cell lines based on their expression levels of GST. To the best of our knowledge, this probe is the first affinity-based iridium(III) imaging probe for the subcellular GST. Our work provides a valuable tool for unmasking the diverse roles of a subcellular GST in living systems, as well as for studying the heterogeneity of breast cancers.

MYC Deregulation and PTEN Loss Model Tumor and Stromal Heterogeneity of Aggressive Triple-Negative Breast Cancer

Triple-negative breast cancer (TNBC) patients have a poor prognosis and few treatment options. Mouse models of TNBC are important for development of new therapies, however, few mouse models represent the complexity of TNBC. Here, we develop a female TNBC murine model by mimicking two common TNBC mutations with high co-occurrence: amplification of the oncogene MYC and deletion of the tumor suppressor PTEN. This Myc;Ptenfl model develops heterogeneous triple-negative mammary tumors that display histological and molecular features commonly found in human TNBC. Our research involves deep molecular and spatial analyses on Myc;Ptenfl tumors including bulk and single-cell RNA-sequencing, and multiplex tissue-imaging. Through comparison with human TNBC, we demonstrate that this genetic mouse model develops mammary tumors with differential survival and therapeutic responses that closely resemble the inter- and intra-tumoral and microenvironmental heterogeneity of human TNBC, providing a pre-clinical tool for assessing the spectrum of patient TNBC biology and drug response.

Development of a CD8+ T cell-based molecular classification for predicting prognosis and heterogeneity in triple-negative breast cancer by integrated analysis of single-cell and bulk RNA-sequencing

BackgroundTriple-negative breast cancer (TNBC), although the most intractable subtype, is characterized by abundant immunogenicity, which enhances responsiveness to immunotherapeutic measures. MethodsFirst, we identified CD8+ T cell core genes (TRCG) based on single-cell sequence and traditional transcriptome sequencing and then used this data to develop a first-of-its-kind classification system based on CD8+ T cells in patients with TNBC. Next, TRCG-related patterns were systematically analyzed, and their correlation with genomic features, immune activity (microenvironment associated with immune infiltration), and clinicopathological characteristics were assessed in the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC), the Cancer Genome Atlas (TCGA), GSE103091, GSE96058 databases. Additionally, a CD8+ T cell-related prognostic signature (TRPS) was developed to quantify a patient-specific TRCG pattern. What's more, the genes-related TRPS was validated by polymerase chain reaction (PCR) experiment. ResultsThis study, for the first time, distinguished two subsets in patients with TNBC based on the TRCG. The immune microenvironment and prognostic stratification between these have distinct heterogeneity. Furthermore, this study constructed a novel scoring system named TRPS, which we show to be a robust prognostic marker for TNBC that is related to the intensity of immune infiltration and immunotherapy. Moreover, the levels of genes related the TRPS were validated by quantitative Real-Time PCR. ConclusionsConsequently, this study unraveled an association between the TRCG and the tumor microenvironment in TNBC. TRPS model represents an effective tool for survival prediction and treatment guidance in TNBC that can also help identify individual variations in TME and stratify patients who are sensitive to anticancer immunotherapy.

Identifying prognostic markers in spatially heterogeneous breast cancer microenvironment

To gain deeper insights into the microenvironment of breast cancer, we utilized GeoMx Digital Spatial Profiling (DSP) technology to analyze transcripts from 107 regions of interest in 65 untreated breast cancer tissue samples. Our study revealed spatial heterogeneity in the expression of marker genes in tumor cell enriched, immune cell enriched, and normal epithelial areas. We evaluated a total of 55 prognostic markers in tumor cell enriched regions and 15 in immune cell enriched regions, identifying that tumor cell enriched regions had higher levels of follicular helper T cells, resting dendritic cells, and plasma cells than immune cell enriched regions, while the levels of resting CD4 memory in T cells and regulatory (Treg) T cells were lower. Additionally, we analyzed the heterogeneity of HLA gene families, immunological checkpoints, and metabolic genes in these areas. Through univariate Cox analysis, we identified 5 prognosis-related metabolic genes. Furthermore, we conducted immunostaining experiments, including EMILIN2, SURF4, and LYPLA1, to verify our findings. Our investigation into the spatial heterogeneity of the breast cancer tumor environment has led to the discovery of specific diagnostic and prognostic markers in breast cancer.

Whole-Exome Sequencing Reveals High Mutational Concordance between Primary and Matched Recurrent Triple-Negative Breast Cancers.

Triple-negative breast cancer (TNBC) is a molecularly complex and heterogeneous breast cancer subtype with distinct biological features and clinical behavior. Although TNBC is associated with an increased risk of metastasis and recurrence, the molecular mechanisms underlying TNBC metastasis remain unclear. We performed whole-exome sequencing (WES) analysis of primary TNBC and paired recurrent tumors to investigate the genetic profile of TNBC. Genomic DNA extracted from 35 formalin-fixed paraffin-embedded tissue samples from 26 TNBC patients was subjected to WES. Of these, 15 were primary tumors that did not have recurrence, and 11 were primary tumors that had recurrence (nine paired primary and recurrent tumors). Tumors were analyzed for single-nucleotide variants and insertions/deletions. The tumor mutational burden (TMB) was 7.6 variants/megabase in primary tumors that recurred (n = 9); 8.2 variants/megabase in corresponding recurrent tumors (n = 9); and 7.3 variants/megabase in primary tumors that did not recur (n = 15). MUC3A was the most frequently mutated gene in all groups. Mutations in MAP3K1 and MUC16 were more common in our dataset. No alterations in PI3KCA were detected in our dataset. We found similar mutational profiles between primary and paired recurrent tumors, suggesting that genomic features may be retained during local recurrence.

Molecular features and clinical implications of the heterogeneity in Chinese patients with HER2-low breast cancer

The molecular heterogeneity and distinct features of HER2-low breast cancers, particularly in the Chinese population, are not well understood, limiting its precise management in the era of antibody‒drug conjugates. To address this issue, we established a cohort of 434 Chinese patients with HER2-low breast cancer (433 female and one male) and integrated genomic, transcriptomic, proteomic, and metabolomic profiling data. In this cohort, HER2-low tumors are more distinguished from HER2-0 tumors in the hormone receptor–negative subgroup. Within HER2-low tumors, significant interpatient heterogeneity also exists in the hormone receptor–negative subgroup: basal-like tumors resemble HER2-0 disease, and non-basal-like HER2-low tumors mimic HER2-positive disease. These non-basal-like HER2-low tumors are enriched in the HER2-enriched subtype and the luminal androgen receptor subtype and feature PIK3CA mutation, FGFR4/PTK6/ERBB4 overexpression and lipid metabolism activation. Among hormone receptor–positive tumors, HER2-low tumors show less loss/deletion in 17q peaks than HER2-0 tumors. In this work, we reveal the heterogeneity of HER2-low breast cancers and emphasize the need for more precise stratification regarding hormone receptor status and molecular subtype.

Breast cancer histopathology image-based gene expression prediction using spatial transcriptomics data and deep learning

Tumour heterogeneity in breast cancer poses challenges in predicting outcome and response to therapy. Spatial transcriptomics technologies may address these challenges, as they provide a wealth of information about gene expression at the cell level, but they are expensive, hindering their use in large-scale clinical oncology studies. Predicting gene expression from hematoxylin and eosin stained histology images provides a more affordable alternative for such studies. Here we present BrST-Net, a deep learning framework for predicting gene expression from histopathology images using spatial transcriptomics data. Using this framework, we trained and evaluated four distinct state-of-the-art deep learning architectures, which include ResNet101, Inception-v3, EfficientNet (with six different variants), and vision transformer (with two different variants), all without utilizing pretrained weights for the prediction of 250 genes. To enhance the generalisation performance of the main network, we introduce an auxiliary network into the framework. Our methodology outperforms previous studies, with 237 genes identified with positive correlation, including 24 genes with a median correlation coefficient greater than 0.50. This is a notable improvement over previous studies, which could predict only 102 genes with positive correlation, with the highest correlation values ranging from 0.29 to 0.34.

Advancing Breast Cancer Heterogeneity Analysis: Insights from Genomics, Transcriptomics and Proteomics at Bulk and Single-Cell Levels.

Breast cancer continues to pose a significant healthcare challenge worldwide for its inherent molecular heterogeneity. This review offers an in-depth assessment of the molecular profiling undertaken to understand this heterogeneity, focusing on multi-omics strategies applied both in traditional bulk and single-cell levels. Genomic investigations have profoundly informed our comprehension of breast cancer, enabling its categorization into six intrinsic molecular subtypes. Beyond genomics, transcriptomics has rendered deeper insights into the gene expression landscape of breast cancer cells. It has also facilitated the formulation of more precise predictive and prognostic models, thereby enriching the field of personalized medicine in breast cancer. The comparison between traditional and single-cell transcriptomics has identified unique gene expression patterns and facilitated the understanding of cell-to-cell variability. Proteomics provides further insights into breast cancer subtypes by illuminating intricate protein expression patterns and their post-translational modifications. The adoption of single-cell proteomics has been instrumental in this regard, revealing the complex dynamics of protein regulation and interaction. Despite these advancements, this review underscores the need for a holistic integration of multiple 'omics' strategies to fully decipher breast cancer heterogeneity. Such integration not only ensures a comprehensive understanding of breast cancer's molecular complexities, but also promotes the development of personalized treatment strategies.