KGANSynergy: knowledge graph attention network for drug synergy prediction.

Combination therapy is widely used to treat complex diseases, particularly in patients who respond poorly to monotherapy. For example, compared with the use of a single drug, drug combinations can reduce drug resistance and improve the efficacy of cancer treatment. Thus, it is vital for researchers and society to help develop effective combination therapies through clinical trials. However, high-throughput synergistic drug combination screening remains challenging and expensive in the large combinational space, where an array of compounds are used. To solve this problem, various computational approaches have been proposed to effectively identify drug combinations by utilizing drug-related biomedical information. In this study, considering the implications of various types of neighbor information of drug entities, we propose a novel end-to-end Knowledge Graph Attention Network to predict drug synergy (KGANSynergy), which utilizes neighbor information of known drugs/cell lines effectively. KGANSynergy uses knowledge graph (KG) hierarchical propagation to find multi-source neighbor nodes for drugs and cell lines. The knowledge graph attention network is designed to distinguish the importance of neighbors in a KG through a multi-attention mechanism and then aggregate the entity's neighbor node information to enrich the entity. Finally, the learned drug and cell line embeddings can be utilized to predict the synergy of drug combinations. Experiments demonstrated that our method outperformed several other competing methods, indicating that our method is effective in identifying drug combinations.

Haplotype-specific MAPK3 expression in 16p11.2 deletion contributes to variable neurodevelopment.

Recurrent proximal 16p11.2 deletion (16p11.2del) is a risk factor for diverse neurodevelopmental disorders with incomplete penetrance and variable expressivity. Although investigation with human induced pluripotent stem cell models has confirmed disruption of neuronal development in 16p11.2del neuronal cells, which genes are responsible for abnormal cellular phenotypes and what determines the penetrance of neurodevelopmental abnormalities are unknown. We performed haplotype phasing of the 16p11.2 region in a 16p11.2del neurodevelopmental disorders cohort and generated human induced pluripotent stem cells for two 16p11.2del families with distinct residual haplotypes and variable neurodevelopmental disorder phenotypes. Using transcriptomic profiles and cellular phenotypes of the human induced pluripotent stem cell-differentiated cortex neuronal cells, we revealed MAPK3 to be a contributor to dysfunction in multiple pathways related to early neuronal development, with altered soma and electrophysiological properties in mature neuronal cells. Notably, MAPK3 expression in 16p11.2del neuronal cells varied on the basis of a 132 kb 58 single nucleotide polymorphism (SNP) residual haplotype, with the version composed entirely of minor alleles associated with reduced MAPK3 expression. Ten SNPs on the residual haplotype were mapped to enhancers of MAPK3. We functionally validated six of these SNPs by luciferase assay, implicating them in the residual haplotype-specific differences in MAPK3 expression via cis-regulation. Finally, the analysis of three different cohorts of 16p11.2del subjects showed that this minor residual haplotype is associated with neurodevelopmental disorder phenotypes in 16p11.2del carriers.

Single-cell analysis of white adipose tissue reveals the tumor-promoting adipocyte subtypes.

BACKGROUND: The tumor-adipose microenvironment (TAME) is characterized by the enrichment of adipocytes, and is considered a special ecosystem that supports cancer progression. However, the heterogeneity and diversity of adipocytes in TAME remains poorly understood. METHODS: We conducted a single-cell RNA sequencing analysis of adipocytes in mouse and human white adipose tissue (WAT). We analyzed several adipocyte subtypes to evaluate their relationship and potential as prognostic factors for overall survival (OS). The potential drugs are screened by using bioinformatics methods. The tumor-promoting effects of a typical adipocyte subtype in breast cancer are validated by performing in vitro functional assays and immunohistochemistry (IHC) in clinical samples. RESULTS: We profiled a comprehensive single-cell atlas of adipocyte in mouse and human WAT and described their characteristics, origins, development, functions and interactions with immune cells. Several cancer-associated adipocyte subtypes, namely DPP4+ adipocytes in visceral adipose and ADIPOQ+ adipocytes in subcutaneous adipose, are identified. We found that high levels of these subtypes are associated with unfavorable outcomes in four typical adipose-associated cancers. Some potential drugs including Trametinib, Selumetinib and Ulixertinib are discovered. Emphatically, knockdown of adiponectin receptor 1 (AdipoR1) and AdipoR2 impaired the proliferation and invasion of breast cancer cells. Patients with AdipoR2-high breast cancer display significantly shorter relapse-free survival (RFS) than those with AdipoR2-low breast cancer. CONCLUSION: Our results provide a novel understanding of TAME at the single-cell level. Based on our findings, several adipocyte subtypes have negative impact on prognosis. These cancer-associated adipocytes may serve as key prognostic predictor and potential targets for treatment in the future.

Phase retrieval via nonlocal complex-domain sparsity.

Phase retrieval is indispensable for a number of coherent imaging systems. Owing to limited exposure, it is a challenge for traditional phase retrieval algorithms to reconstruct fine details in the presence of noise. In this Letter, we report an iterative framework for noise-robust phase retrieval with high fidelity. In the framework, we investigate nonlocal structural sparsity in the complex domain by low-rank regularization, which effectively suppresses artifacts caused by measurement noise. The joint optimization of sparsity regularization and data fidelity with forward models enables satisfying detail recovery. To further improve computational efficiency, we develop an adaptive iteration strategy that automatically adjusts matching frequency. The effectiveness of the reported technique has been validated for coherent diffraction imaging and Fourier ptychography, with ≈7 dB higher peak SNR (PSNR) on average, compared with conventional alternating projection reconstruction.

Erratum: Trehalose inhibits H2O2-induced autophagic death in dopaminergic SH-SY5Y cells via mitigation of ROS-dependent endoplasmic reticulum stress and AMPK activation: Erratum.

[This corrects the article DOI: 10.7150/ijms.25656.].

Efficient large-scale single-pixel imaging.

The speed of single-pixel imaging (SPI) is tied to its resolution, which is positively related to the number of modulation times. Therefore, efficient large-scale SPI is a serious challenge that impedes its wide applications. In this work, we report a novel, to the best of our knowledge, sparse SPI scheme and corresponding reconstruction algorithm to image target scenes at above 1 K resolution with reduced measurements. Specifically, we first analyze the statistical importance ranking of Fourier coefficients for natural images. Then the sparse sampling with a polynomially decending probability of the ranking is performed to cover a larger range of the Fourier spectrum than non-sparse sampling. The optimal sampling strategy with suitable sparsity is summarized for the best performance. Next, a lightweight deep distribution optimization (D2O) algorithm is introduced for large-scale SPI reconstruction from sparsely sampled measurements instead of a conventional inverse Fourier transform (IFT). The D2O algorithm empowers robustly recovering sharp scenes at 1 K resolution within 2 s. A series of experiments demonstrate the technique's superior accuracy and efficiency.

Ultrahigh-security single-pixel semantic encryption.

Single-pixel encryption is a recently developed encryption technique enabling the ciphertext amount to be decreased. It adopts modulation patterns as secret keys and uses reconstruction algorithms for image recovery in the decryption process, which are time-consuming and can easily be illegally deciphered if the patterns are exposed. We report an image-free single-pixel semantic encryption technique that significantly enhances security. The technique extracts semantic information directly from the ciphertext without image reconstruction, which significantly reduces computing resources for end-to-end real-time decoding. Moreover, we introduce a stochastic mismatch between keys and ciphertext, with random measurement shift and dropout, which effectively enhances the difficulty of illegal deciphering. Experiments on the MNIST dataset validate that 78 coupling measurements (0.1 sampling rate) with stochastic shift and random dropout achieved 97.43% semantic decryption accuracy. In the worst situation, when all the keys are illegally obtained by unauthorized attackers, only 10.80% accuracy can be achieved (39.47% in an ergodic manner).

CNV profiles of Chinese pediatric patients with developmental disorders.

PURPOSE: To examine the overall genomic copy-number variant (CNV) landscape of Chinese pediatric patients with developmental disorders. METHODS: De-identified chromosomal microarray (CMA) data from 10,026 pediatric patients with developmental disorders were collected for re-evaluating the pathogenic CNV (pCNV) yields of different medical conditions and for comparing the frequency and phenotypic variability of genomic disorders between the Chinese and Western patient populations. RESULTS: The overall yield of pCNVs in the Chinese pediatric patient cohort was 21.37%, with variable yields for different disorders. Yields of pCNVs were positively associated with phenotypic complexity and intellectual disability/developmental delay (ID/DD) comorbidity for most disorders. The genomic burden and pCNV yield in neurodevelopmental disorders supported a female protective effect. However, the stratification analysis revealed that it was seen only in nonsyndromic ID/DD, not in nonsyndromic autism spectrum disorders or seizure. Furthermore, 15 known genomic disorders showed significantly different frequencies in Chinese and Western patient cohorts, and profiles of referred clinical features for 15 known genomic disorders were also significantly different in the two cohorts. CONCLUSION: We defined the pCNV yields and profiles of the Chinese pediatric patients with different medical conditions and uncovered differences in the frequency and phenotypic diversity of genomic disorders between Chinese and Western patients.

Short-term and long-term safety and efficacy of tenofovir alafenamide, tenofovir disoproxil fumarate and entecavir treatment of acute-on-chronic liver failure associated with hepatitis B.

BACKGROUND & AIMS: There is limited evidence on the efficacy and safety of nucleos(t) ide analogues (NAs) in the treatment of HBV-ACLF. Our objective was to evaluate the outcomes among TAF, TDF and ETV, three first-line antivirals against chronic hepatitis B, in patients with HBV-ACLF. METHODS: Patients with HBV-related ACLF were recruited and received daily TAF (25 mg/d), TDF (300 mg/d) and ETV (0.5 mg/d). They were prospectively followed-up. The primary endpoint was overall survival at week 12 and week 48, the secondary endpoints were virological response and biochemical response. RESULTS: Forty gender and age matched eligible subjects were recruited and divided into three groups: TAF group, TDF group and ETV group. By week 48, 8 (80%) patients in TAF group, 6 (60%) patients in TDF group and 17 (85%) patients in ETV group survived without liver transplantation (P = 0.251). After 4 weeks of NAs treatment, all three groups showed paralleling reduction of HBV DNA levels. All three groups presented similar biochemical responses at week 4, patients treated with TAF showed a priority in total bilirubin reduction, albumin and cholesterol maintenance. Additionally, although there was no significant difference in changes of serum urea, serum creatinine, serum cystatin C and estimated GFR among the three groups by treatment week 4, TDF showed unfavorable renal safety even in short -term treatment. The treatment using NAs was well-tolerated and there was no serious drug-related adverse event reported. CONCLUSIONS: TAF, TDF and ETV are of similar efficacy and safety in short-term and long-term treatment of HBV-ACLF. TRIAL REGISTRATION: This study is ongoing and is registered with ClinicalTrials.gov , NCT03640728 (05/02/2019).

[Autism spectrum disorder/development delay in siblings with SCN2A mutations caused by germline mosaicism].

OBJECTIVE: To report on a family which has two siblings with SCN2A mutation caused by germline mosaicism suffering from autism spectrum disorder/development delay (ASD/DD). METHODS: Clinical data was collected for the proband and his parents. Next generation sequencing (NGS) was carried out on the proband and his parents. Suspected mutations were verified by Sanger sequencing of the proband, his parents and brother. To detect whether there is a low proportion of somatic mosaicism in the parents, a droplet digital PCR was conducted. The result of ddPCR showed that the father was germline mosaicism (0.233%). RESULTS: NGS has identified a de novo splicing mutation of the SCN2A gene, c.605+1G>A, in the proband and his brother. Combined with its clinical phenotype and inheritance pattern, SCN2A was judged to be the pathogenic gene. Above findings strongly suggested parental germline mosaicism. CONCLUSION: ASD/DD in siblings with SCN2A mutations caused by germline mosaicism. Paternal mosaicism should be considered as one of the important inheritance patterns for counseling parents with a child carrying SCN2A mutation. The ddPCR can help to reveal very low proportion of germline mosaicism.

[Clinical features and gene variant of a pedigree affected with X-linked recessive mental retardation Claes-Jensen type].

OBJECTIVE: To explore the genetic basis for a pedigree affected with X-linked recessive mental retardation Claes-Jensen type. METHODS: Genomic DNA was extracted from peripheral blood samples of the patient, his parents (phenotypically normal) and two elder brothers with similar clinical manifestations. Whole exome sequencing was carried out for the proband, and the result was verified by Sanger sequencing. RESULTS: The proband was found to harbor a hemizygous c.1565C>T missense variant in exon 11 of the KDM5C gene. The transition has resulted in replacement of serine by phenylalanine at position 522 (p.Ser522Phe). Sanger sequencing showed that the patient's two elder brothers and mother carried the same variant, which was predicted to be probably damaging by SIFT, PolyPhen2 and Mutation_Taster. The three affected brothers presented with similar clinical phenotypes characterized by mental retardation, speech delay, behavioral problem, self-limited epilepsy responsible to medication, short stature and microcephaly. The mother only had mild cognitive impairment and learning disability. The same variant was not found in their father and was unreported previously. CONCLUSION: The c.1565C>T (p.Ser522Phe) of the KDM5C gene probably underlay the X-linked recessive mental retardation Claes-Jensen type in this pedigree.

Identification of two novel null variants in CLN8 by targeted next-generation sequencing: first report of a Chinese patient with neuronal ceroid lipofuscinosis due to CLN8 variants.

BACKGROUND: Neuronal ceroid lipofuscinoses (NCLs) are one of the most frequent childhood-onset neurodegenerative pathologies characterized by seizures, progressive cognitive decline, motor impairment and loss of vision. For the past two decades, more than 430 variants in 13 candidate genes have been identified in the affected patients. Most of the variants were almost exclusively reported in Western patients, and very little clinical and genetic information was available for Chinese patients. CASE PRESENTATION: We report a Chinese boy whose clinical phenotypes were suspected to be NCL, including intractable epilepsy, cognitive and motor decline and progressive vision loss. Using targeted next-generation sequencing, two novel null variants in CLN8 (c.298C > T, p.Gln100Ter; c.551G > A, p.Trp184Ter) were detected in this patient in trans model. These two variants were interpreted as pathogenic according to the variant guidelines of the American College of Medical Genetics and Genomics. CONCLUSIONS: This is the first case report of NCL due to CLN8 variants in China. Our findings expand the variant diversity of CLN8 and demonstrate the tremendous diagnosis value of targeted next-generation sequencing for pediatric NCLs.

Trehalose inhibits H2O2-induced autophagic death in dopaminergic SH-SY5Y cells via mitigation of ROS-dependent endoplasmic reticulum stress and AMPK activation.

Autophagy is a catabolic process to maintain intracellular homeostasis via removal of cytoplasmic macromolecules and damaged cellular organelles through lysosome-mediated degradation. Trehalose is often regarded as an autophagy inducer, but we reported previously that it could prevent ischemic insults-induced autophagic death in neurons. Thus, we further investigated in this study whether trehalose could protect human dopaminergic SH-SY5Y cells against H2O2-induced lethal autophagy. We found pretreatment with trehalose not only prevented H2O2-induced death in SH-SY5Y cells, but also reversed H2O2-induced upregulation of LC3II, Beclin1 and ATG5 and downregulation of p62. Then, we proved that either autophagy inhibitor 3MA or genetic knockdown of ATG5 prevented H2O2-triggered death in SH-SY5Y cells. These indicated that trehalose could inhibit H2O2-induced autophagic death in SH-SY5Y cells. Further, we found that trehalose inhibited H2O2-induced AMPK activation and endoplasmic reticulum (ER) stress. Moreover, inhibition of AMPK activation with compound C or alleviation of ER stress with chemical chaperone 4-PBA obviously attenuated H2O2-induced changes in autophagy-related proteins. Notably, we found that trehalose inhibited H2O2-induced increase of intracellular ROS and reduction in the activities of CAT and SOD. Consistently, our data revealed as well that mitigation of intracellular ROS levels with antioxidant NAC markedly attenuated H2O2-induced AMPK activation and ER stress. Therefore, we demonstrated in this study that trehalose prevented H2O2-induced autophagic death in SH-SY5Y cells via mitigation of ROS-dependent endoplasmic reticulum stress and AMPK activation.

[Clinical and genetic analysis of a patient with cutis laxa].

OBJECTIVE To identify potential mutation in a patient with cutis laxa through exome sequencing of genetic disease-related genes and explore its clinical and genetic features. METHODS Clinical data was collected for the proband and her parents. Exome sequencing was carried out on the proband. Suspected mutations were verified by Sanger sequencing. RESULTS Exome sequencing identified a compound heterozygous mutation of the ATP6V0A2 gene, c.187C>T (p.R63X) and c.1189G>C (p.A397P), in the proband. The mutations were respectively inherited from the father and mother. The patient was diagnosed with autosomal recessive cutis laxa type 2A (ARCL2A). CONCLUSION A case with ARCL2A was diagnosed. The novel mutation has expanded the spectrum of ATP6V0A2 mutations. Exome sequencing is a useful tool for the diagnosis of complex genetic diseases.

[A clinical and genetic analysis of a child with supernumerary marker chromosome 15-caused mental retardation, intractable epilepsy, and central precocious puberty].

Supernumerary marker chromosome 15 is a rare chromosome abnormality. This paper reports the clinical diagnosis and treatment, as well as genetic defects, of a child with supernumerary marker chromosome 15. The patient was a 9.5-year-old girl who had mental and motor retardation since infancy, breast development at the age of 7 years, and seizures at the age of 8.5 years. Seizures occurred with various features and could not be controlled by a variety of antiepileptic drugs. No abnormalities were found by brain magnetic resonance imaging. Electroencephalogram showed frequent epileptiform discharges. G-banding karyotype analysis, fluorescence in situ hybridization, methylation-specific multiplex ligation-dependent probe amplification, and array comparative genomic hybridization identified a de novo 15q duplication in the patient. The maternal copy number increased in the 15q11-13 region. The form of genome rearrangement was 47,XX,+inv dup(15)(pter to q13:q13 to pter). The increased copy number in the 15q11-13 region is closely related to mental retardation, intractable epilepsy, and central precocious puberty. High-resolution karyotype analysis is recommended for children with unexplained mental retardation and epilepsy.

[Glucose transporter 1 deficiency syndrome: features of movement disorders, diagnosis and treatment].

OBJECTIVE: To investigate the clinical features, diagnosis and treatment of glucose transporter 1 deficiency syndrome (GLUT1-DS), as well as the diagnostic value of movement disorders. METHODS: The clinical data of four children with GLUT1-DS were collected, and their clinical features, treatment, and follow-up results were analyzed. RESULTS: There were two boys and two girls, with an age of onset of 2-15 months. Clinical manifestations included movement disorders, seizures, and developmental retardation. Seizures were the cause of the first consultation in all cases. The four children all had persistent ataxia, dystonia, and dysarthria; two had persistent tremor, two had paroxysmal limb paralysis, and two had eye movement disorders. Paroxysmal symptoms tended to occur in fatigue state. All four children had reductions in the level of cerebrospinal fluid glucose and its ratio to blood glucose, as well as SLC2A1 gene mutations. The four children were given a ketogenic diet, at a ketogenic ratio of 2:1 to 3:1, and achieved complete remission of paroxysmal symptoms within 5 weeks. CONCLUSIONS: GLUT1-DS should be considered for epileptic children with mental retardation and motor developmental delay complicated by various types of movement disorders. The ketogenic diet is effective at a ketogenic ratio of 2:1 to 3:1 for the treatment of GLUT1-DS.

[Clinical and molecular genetic study of nonketotic hyperglycinemia in a Chinese family].

Nonketotic hyperglycinemia (NKH) is a rare, inborn error of metabolism. In this case report, a Chinese male infant was diagnosed with NKH caused by GLDC gene mutation. The clinical characteristics and genetic diagnosis were reported. The infant presented with an onset of early metabolic encephalopathy and Ohtahara syndrome. Both blood and urinary levels of metabolites were in the normal range. Brain MRI images indicated a poor development of corpus callosum, and a burst suppression pattern was found in the EEG. Results of target gene sequencing technology combined with multiplex ligation-dependent probe amplification (MLPA) indicated a heterozygous missense mutation of c.1786 C>T (p.R596X) in maternal exon 15 and a loss of heterozygosity of 4-15 exon gross deletions in paternal GLDC gene. These definite pathogenic mutations confirmed the diagnosis of NKH. The infant's clinical condition was not improved after treatment with adreno-cortico-tropic-hormone, topiramate and dextromethorphan, and he finally died at 4 months of age. Patients with NKH often exhibit complicated clinical phenotypes and are lack of specific symptoms. NKH could be diagnosed by metabolic screening and molecular genetic analysis.

Cancer-associated fibroblast-derived gene signature discriminates distinct prognoses by integrated single-cell and bulk RNA-seq analyses in breast cancer.

BACKGROUND: Cancer-associated fibroblasts (CAFs) are one of the most predominant cellular subpopulations in the tumor stroma and play an integral role in cancer occurrence and progression. However, the prognostic role of CAFs in breast cancer remains poorly understood. METHODS: We identified a number of CAF-related biomarkers in breast cancer by combining single-cell and bulk RNA-seq analyses. Based on univariate Cox regression as well as Least Absolute Shrinkage and Selection Operator (LASSO) regression analysis, a novel CAF-associated prognostic model was developed. Breast cancer patients were grouped according to the median risk score and further analyzed for outcome, clinical characteristic, pathway activity, genomic feature, immune landscape, and drug sensitivity. RESULTS: A total of 341 CAF-related biomarkers were identified from single-cell and bulk RNA-seq analyses. We eventually screened eight candidate prognostic genes, including CERCAM, EMP1, SDC1, PRKG1, XG, TNN, WLS, and PDLIM4, and constructed the novel CAF-related prognostic model. Grouped by the median risk score, high-risk patients showed a significantly worse prognosis and exhibited distinct pathway activities such as uncontrolled cell cycle progression, angiogenesis, and activation of glycolysis. In addition, the combined risk score and tumor mutation burden significantly improved the ability to predict patient prognosis. Importantly, patients in the high-risk group had a higher infiltration of M2 macrophages and a lower infiltration of CD8+ T cells and activated NK cells. Finally, we calculated the IC50 for a range of anticancer drugs and personalized the treatment regimen for each patient. CONCLUSION: Integrating single-cell and bulk RNA-seq analyses, we identified a list of compositive CAF-associated biomarkers and developed a novel CAF-related prognostic model for breast cancer. This robust CAF-derived gene signature acts as an excellent predictor of patient outcomes and treatment responses in breast cancer.

Comprehensive single-cell and bulk RNA-seq analyses reveal a novel CD8+ T cell-associated prognostic signature in ovarian cancer.

CD8+ T cells play pivotal roles in combating intracellular pathogens and eliminating malignant cells in cancer. However, the prognostic role of CD8+ T cells in ovarian carcinoma is insufficiently exploited. Herein, through univariate Cox regression along with least absolute shrinkage and selection operator (LASSO) regression analyses, we developed a novel prognostic model based on CD8+ T cell markers identified by single-cell sequencing (scRNA-seq) analyses. Patient grouping by the median risk score reveals an excellent prognostic efficacy of this model in both training and validation cohorts. Of note, patients classified as low-risk group exhibit a dramatically improved prognosis. In addition, higher enrichment level of immune-related pathways and increased infiltration level of multiple immune cells are found in patients with lower risk score. Importantly, low-risk patients also exhibited higher response rate to immunotherapies. Summarily, this developed CD8+ T cell-associated prognostic model serves as an excellent predictor for clinical outcomes and aids in guiding therapeutic strategy choices for ovarian cancer patients.

Identification and immunological role of cuproptosis in osteoporosis.

Background: osteoporosis is a skeletal disorder disease features low bone mass and poor bone architecture, which predisposes to increased risk of fracture. Copper death is a newly recognized form of cell death caused by excess copper ions, which presumably involve in various disease. Accordingly, we intended to investigate the molecular clusters related to the cuproptosis in osteoporosis and to construct a predictive model. Methods: we investigated the expression patterns of cuproptosis regulators and immune signatures in osteoporosis based on the GSE56815 dataset. Through analysis of 40 osteoporosis samples, we investigated molecular clustering on the basis of cuproptosis--related genes, together with the associated immune cell infiltration. The WGCNA algorithm was applied to detect cluster-specific differentially expressed genes. Afterwards, the optimum machine model was selected by calculating the performance of the support vector machine model, random forest model, eXtreme Gradient Boosting and generalized linear model. Nomogram, decision curve analysis, calibration curves, and the GSE7158 dataset was utilizing to confirm the prediction efficiency. Results: Differences between osteoporotic and non-osteoporotic controls confirm poorly adjusted copper death-related genes and triggered immune responses. In osteoporosis, two clusters of molecules in connection with copper death proliferation were outlined. The assessed levels of immune infiltration showed prominent heterogeneity between the different clusters. Cluster 2 was characterized by a raised immune score accompanied with relatively high levels of immune infiltration. The functional analysis we performed showed a close relationship between the different immune responses and specific differentially expressed genes in cluster 2. The random forest machine model showed the optimum discriminatory performance due to relatively low residuals and root mean square errors. Finally, a random forest model based on 5 genes was built, showing acceptable performance in an external validation dataset (AUC = 0.750). Calibration curve, Nomogram, and decision curve analyses also evinced fidelity in predicting subtypes of osteoporosis. Conclusion: Our study identifies the role of cuproptosis in OP and essentially illustrates the underlying molecular mechanisms that lead to OP heterogeneity.