HyperTMO: a trusted multi-omics integration framework based on hypergraph convolutional network for patient classification.

MOTIVATION: The rapid development of high-throughput biomedical technologies can provide researchers with detailed multi-omics data. The multi-omics integrated analysis approach based on machine learning contributes a more comprehensive perspective to human disease research. However, there are still significant challenges in representing single-omics data and integrating multi-omics information. RESULTS: This article presents HyperTMO, a Trusted Multi-Omics integration framework based on Hypergraph convolutional network for patient classification. HyperTMO constructs hypergraph structures to represent the association between samples in single-omics data, then evidence extraction is performed by hypergraph convolutional network, and multi-omics information is integrated at an evidence level. Last, we experimentally demonstrate that HyperTMO outperforms other state-of-the-art methods in breast cancer subtype classification and Alzheimer's disease classification tasks using multi-omics data from TCGA (BRCA) and ROSMAP datasets. Importantly, HyperTMO is the first attempt to integrate hypergraph structure, evidence theory, and multi-omics integration for patient classification. Its accurate and robust properties bring great potential for applications in clinical diagnosis. AVAILABILITY AND IMPLEMENTATION: HyperTMO and datasets are publicly available at https://github.com/ippousyuga/HyperTMO.

Large-Area Near-Infrared Emission Enhancement on Single Upconversion Nanoparticles by Metal Nanohole Array.

Single lanthanide (Ln) ion doped upconversion nanoparticles (UCNPs) exhibit great potential for biomolecule sensing and counting. Plasmonic structures can improve the emission efficiency of single UCNPs by modulating the energy transferring process. Yet, achieving robust and large-area single UCNP emission modulation remains a challenge, which obstructs investigation and application of single UCNPs. Here, we present a strategy using metal nanohole arrays (NHAs) to achieve energy-transfer modulation on single UCNPs simultaneously within large-area plasmonic structures. By coupling surface plasmon polaritons (SPPs) with higher-intermediate state (1D2 → 3F3, 1D2 → 3H4) transitions, we achieved a remarkable up to 10-fold enhancement in 800 nm emission, surpassing the conventional approach of coupling SPPs with an intermediate ground state (3H4 → 3H6). We numerically simulate the electrical field distribution and reveal that luminescent enhancement is robust and insensitive to the exact location of particles. It is anticipated that the strategy provides a platform for widely exploring applications in single-particle quantitative biosensing.

Single-cell RNA-seq reveals a microenvironment and an exhaustion state of T/NK cells in acute myeloid leukemia.

Acute myeloid leukemia (AML) is a heterogeneous blood cancer. Effective immunotherapies for AML are hindered by a lack of understanding of the tumor microenvironment (TME). Here, we retrieved published single-cell RNA sequencing data for 128,688 cells derived from 29 bone marrow aspirates, including 21 AML patients and eight healthy donors. We established a global tumor ecosystem including nine main cell types. Myeloid, T, and NK cells were further re-clustered and annotated. Developmental trajectory analysis indicated that exhausted CD8+ T cells might develop via tissue residual memory T cells (TRM) in the AML TME. Significantly higher expression levels of exhaustion molecules in AML TRM cells suggested that these cells were influenced by the TME and entered an exhausted state. Meanwhile, the upregulation of checkpoint molecules and downregulation of granzyme were also observed in AML NK cells, suggesting an exhaustion state. In conclusion, our comprehensive profiling of T/NK subpopulations provides deeper insights into the AML immunosuppressive ecosystem, which is critical for immunotherapies.

CircIQGAP1 regulates RCAN1 and RCAN2 through the mechanism of ceRNA and promotes the growth of malignant glioma.

Circular RNA (circRNA) is one of non-coding RNA with specific circular structure, which has been found to be involved in regulating a series of malignant biological behaviors in many malignant tumors. In this study, based on the IDH1 molecular typing of gliomas, we identified a significant downregulation of circRNA (circIQGAP1) expression in IDH1 mutant gliomas by high-throughput sequencing. In 79 tissue samples, we confirmed that circIQGAP1 expression was significantly downregulated in IDH1 mutant gliomas, and that low circIQGAP1 expression was positively associated with better prognosis. Knockdown of circIQGAP1 in glioma cell lines inhibited glioma cell malignancy and conversely overexpression of circIQGAP1 promoted glioma malignancy. circIQGAP1 regulated glioma cell migration, proliferation, invasion and apoptosis through miR-1256/RCAN1/Bax/Bcl-2/Caspase3 and miR-622/RCAN2/Bax/Bcl-2/Caspase3 axes. These results suggest that circIQGAP1 plays an important role in glioma development, promotes tumor growth, and is a potential therapeutic target for glioma.

Synthesis, Crystal Structure, and Nonlinear-Optical Properties of a Diamond-Like Chalcogenide Cu2GeS3.

Metal sulfides with diamond-like (DL) structures generally exhibit excellent mid-IR nonlinear-optical (NLO) properties. Here, Cu2GeS3 (CGS) as a member of the DL chalcogenides was synthesized by a high-temperature solid-state method, and the optical properties were carefully studied experimentally and theoretically. The results revealed that CGS has a large second harmonic generation (0.8 × AgGaSe2) and a moderate birefringence of 0.067 at 1064 nm. In addition, the linear and NLO properties of the A2MS3 (A = Cu, Li; M = Ge, Si) series of compounds were evaluated and compared with the help of first-principles calculations.

Prediction of 3-year risk of diabetic kidney disease using machine learning based on electronic medical records.

BACKGROUND: Established prediction models of Diabetic kidney disease (DKD) are limited to the analysis of clinical research data or general population data and do not consider hospital visits. Construct a 3-year diabetic kidney disease risk prediction model in patients with type 2 diabetes mellitus (T2DM) using machine learning, based on electronic medical records (EMR). METHODS: Data from 816 patients (585 males) with T2DM and 3 years of follow-up at the PLA General Hospital. 46 medical characteristics that are readily available from EMR were used to develop prediction models based on seven machine learning algorithms (light gradient boosting machine [LightGBM], eXtreme gradient boosting, adaptive boosting, artificial neural network, decision tree, support vector machine, logistic regression). Model performance was evaluated using the area under the receiver operating characteristic curve (AUC). Shapley additive explanation (SHAP) was used to interpret the results of the best performing model. RESULTS: The LightGBM model had the highest AUC (0.815, 95% CI 0.747-0.882). Recursive feature elimination with random forest and SHAP plot based on LightGBM showed that older patients with T2DM with high homocysteine (Hcy), poor glycemic control, low serum albumin (ALB), low estimated glomerular filtration rate (eGFR), and high bicarbonate had an increased risk of developing DKD over the next 3 years. CONCLUSIONS: This study constructed a 3-year DKD risk prediction model in patients with T2DM and normo-albuminuria using machine learning and EMR. The LightGBM model is a tool with potential to facilitate population management strategies for T2DM care in the EMR era.

A design method based on Bayesian decision for routing-based digital microfluidic biochips.

With the continuous application and development of the digital microfluidic technology in various fields, many researchers have studied the design of digital microfluidic chips. Module-based chip design methods greatly simplify the design process but waste resources, including through the inadequate use of electrodes within the module and guard cells. To address this problem, a routing-based synthesis method based on a digital microfluidic biochip (DMFB) platform is presented. Routing-based DMFBs ensure a much higher chip utilization factor by removing the virtual modules on the chip and the extra electrodes needed as guard cells. Many previous works focused only on the problems of synthesis completion times, bioassay completion times, and electrode utilization rates. However, the reliability of chips has not been fully studied, and this factor is extremely important because faulty chips affect the test results. Thus, the influence of chip reliability should be fully considered. This paper proposes a design method based on Bayesian decision-making (BBD) for routing-based DMFBs that can fully consider the reliability of chips during the DMFB design process. Simulated experimental results showed that the method can address the reliability problems of chips. The efficiency and convergence performance of the algorithm were very good. The proposed method can achieve an average assay completion time that is shorter than those of the moduleless synthesis (MLS) and modified-MLS (MMLS) methods. The electrode usage rate of the proposed method is better than that of the module-based and improved Dijkstra and improved particle swarm optimization (ID-IPSO) methods.

Down-conversion luminescence and temperature sensing characteristics of LaSrGaO4 :Er3.

Erbium(III) ion (Er3+ ) has abundant energy levels that can emit light covering a quite broad wavelength range in many hosts. Here we synthesized LaSrGaO4 :Er3+ phosphors by a high-temperature solid-state method. Upon excitation at the ultraviolet (UV) band, LaSrGaO4 :Er3+ phosphors could emit green, red and near-infrared emission simultaneously. The temperature dependent emission characteristics of the as-prepared samples was then studied and two kinds of luminescent ratiometric thermometry were constructed. The first one is on the basis of two green emission bands that stems from the 2 H11/2 → 4 I15/2 and 4 S3/2 → 4 I15/2 transitions of Er3+ . The intensity ratio between these two emission bands was found to follow well with the Boltzmann distribution, and its maximum relative sensitivity was calculated to be 0.84% K-1 at 299 K. The other one depends on the 4 F9/2 → 4 I15/2 transition of Er3+ and self-luminescence of the host LaSrGaO4 , considering that these two emission lines have different temperature response. The relative sensitivity of this type of luminescence intensity ratio (LIR) thermometry could reach 1.86% K-1 at 299 K, we have successfully developed materials with one of the largest relative sensitivities to date, which provides some basis for the subsequent development of a new type of non-contact temperature sensor.

On learning disentangled representations for individual treatment effect estimation.

OBJECTIVE: Estimating the individualized treatment effect (ITE) from observational data is a challenging task due to selection bias, which results from the distributional discrepancy between different treatment groups caused by the dependence between features and assigned treatments. This dependence is induced by the factors related to the treatment assignment. We hypothesize that features consist of three types of latent factors: outcome-specific factors, treatment-specific factors and confounders. Then, we aim to reduce the influence of treatment-related factors, i.e., treatment-specific factors and confounders, on outcome prediction to mitigate the effects of selection bias. METHOD: We present a novel representation learning model in which both the main task of outcome prediction and the auxiliary task of classifying the treatment assignment are used to learn the outcome-oriented and treatment-oriented latent representations, respectively. However, since the confounders are related to both treatment assignment and outcome, it is still contained in the representations. To further reduce influence of the confounders contained in both representations, individualized orthogonal regularization is incorporated into the proposed model. The orthogonal regularization forces the outcome-oriented and treatment-oriented latent representations of an individual to be vertical in the inner product space, meaning they are orthogonal with each other, and the common information of confounder is reduced. Such that the ITE can be estimated more precisely without the effects of selection bias. RESULT: We evaluate our proposed model on a semi-simulated dataset and a real-world dataset. The experimental results demonstrate that the proposed model achieves competitive or better performance compared with the performances of the state-of-the-art models. CONCLUSION: The proposed method is well performed on ITE estimation with the ability to reduce selection bias thoroughly by incorporating an auxiliary task and adopting orthogonal regularization to disentangle the latent factors. SIGNIFICANCE: This paper offers a novel method of reducing selection bias in estimating the ITE from observational data by disentangled representation learning.

Knowledge-aware multi-center clinical dataset adaptation: Problem, method, and application.

Adaptable utilization of clinical data collected from multiple centers, prompted by the need to overcome the shifts between the dataset distributions, and exploit these different datasets for potential clinical applications, has received significant attention in recent years. In this study, we propose a novel approach to this task by infusing an external knowledge graph (KG) into multi-center clinical data mining. Specifically, we propose an adversarial learning model to capture shared patient feature representations from multi-center heterogeneous clinical datasets, and employ an external KG to enrich the semantics of the patient sample by providing both clinical center-specific and center-general knowledge features, which are trained with a graph convolutional autoencoder. We evaluate the proposed model on a real clinical dataset extracted from the general cardiology wards of a Chinese hospital and a well-known public clinical dataset (MIMIC III, pertaining to ICU clinical settings) for the task of predicting acute kidney injury in patients with heart failure. The achieved experimental results demonstrate the efficacy of our proposed model.

Up-regulation of DDIT4 predicts poor prognosis in acute myeloid leukaemia.

The mammalian target of rapamycin (mTOR) inhibitor, DNA damage inducible transcript 4 (DDIT4), has inducible expression in response to various cellular stresses. In multiple malignancies, studies have shown that DDIT4 participates in tumorigenesis and impacts patient survival. We aimed to study the prognostic value of DDIT4 in acute myeloid leukaemia (AML), which is currently unclear. Firstly, The Cancer Genome Atlas was screened for AML patients with complete clinical characteristics and DDIT4 expression data. A total of 155 patients were included and stratified according to the treatment modality and the median DDIT4 expression levels. High DDIT4 expressers had shorter overall survival (OS) and event-free survival (EFS) than the low expressers among the chemotherapy-only group (all P < .001); EFS and OS were similar in the high and low DDIT4 expressers of the allogeneic haematopoietic stem cell transplantation (allo-HSCT) group. Furthermore, in the DDIT4high group, patients treated with allo-HSCT had longer EFS and OS than those who received chemotherapy alone (all P < .01). In the DDIT4low group, OS and EFS were similar in different treatment groups. Secondly, we analysed two other cytogenetically normal AML (CN-AML) cohorts derived from the Gene Expression Omnibus database, which confirmed that high DDIT4 expression was associated with poorer survival. Gene Ontology (GO) enrichment analysis showed that the genes related to DDIT4 expression were mainly concentrated in the acute and chronic myeloid leukaemia signalling pathways. Collectively, our study indicates that high DDIT4 expression may serve as a poor prognostic factor for AML, but its prognostic effects could be outweighed by allo-HSCT.

Prognostic role of SCAMP family in acute myeloid leukemia.

Acute myeloid leukemia (AML) is a malignant disease of myeloid hematopoietic stem or progenitor cells characterized by abnormal proliferation of primary and immature myeloid cells in bone marrow and peripheral blood. Gene mutation and expression profiles can be used as prognosis predictors for different prognostic subgroups. Secretory carrier-associated membrane proteins (SCAMPs) are a multigenic family with five members and act as cell surface vectors in the post-Golgi recycling pathways in mammals. Nevertheless, the prognostic and clinical influence of SCAMP family has hardly ever been illustrated in AML. In our study, expression patterns of SCAMP family (SCAMP1-5) were analyzed in 155 AML patients which were extracted from the Cancer Genome Atlas database. In chemotherapy, only subgroup, higher SCAMP1 level was significantly associated with longer EFS and OS (all P = 0.002), and SCAMP1 was confirmed to be an independent favorable factor in un-transplanted patients by Multivariate analysis (all P < 0.05). Nevertheless, in the allogeneic hematopoietic stem cell transplantation (allo-HSCT) treatment subgroup, none of the SCAMP genes had any effect on the clinical survival. Our study found that high expression level of SCAMP1 is a favorable prognostic factor in AML, but allo-HSCT may neutralize its prognostic effect.

High expression of chaperonin-containing TCP1 subunit 3 may induce dismal prognosis in multiple myeloma.

The prognosis role of CCT3 in MM and the possible pathways it involved were studied in our research. By analyzing ten independent datasets (including 48 healthy donors, 2220 MM, 73 MGUS, and 6 PCL), CCT3 was found to express higher in MM than healthy donors, and the expression level was gradually increased from MGUS, SMM, MM to PCL (all P < 0.01). By analyzing three independent datasets (GSE24080, GSE2658, and GSE4204), we found that CCT3 was a significant indicator of poor prognosis (all P < 0.01). KEGG and GSEA analysis showed that CCT3 expression was associated with JAK-STAT3 pathway, Hippo signaling pathway, and WNT signaling pathway. In addition, different expressed genes analysis revealed MYC, which was one of the downstream genes regulated by JAK-STAT3 pathway, was upregulated in MM. This confirms that JAK-STAT3 signaling pathway may promote the progress of disease which was regulated by CCT3 expression. Our study revealed that CCT3 may play a supporting role at the diagnosis of myeloid, and high expression of CCT3 suggested poor prognosis in MM. CCT3 expression may promote the progression of MM mainly by regulating MYC through JAK-STAT3 signaling pathway.

Detecting Rare Variants and Heteroplasmy of Mitochondrial DNA from High-Throughput Sequencing in Patients with Coronary Artery Disease.

BACKGROUND Although mutations and dysfunction of mitochondrial DNA (mtDNA) are related to a variety of diseases, few studies have focused on the relationship between mtDNA and coronary artery disease (CAD), especially the relationship between rare variants and CAD. MATERIAL AND METHODS Two-stage high-throughput sequencing was performed to detect mtDNA variants or heteroplasmy and the relationship between them and CAD phenotypes. In the discovery stage, mtDNA was analyzed by high-throughput sequencing of long-range PCR products generated from the peripheral blood of 85 CAD patients and 80 demographically matched controls. In the validation stage, high-throughput sequencing for mtDNA target regions captured by GenCap Kit was performed on 100 CAD samples and 100 controls. Finally, tRNA fine mapping was performed between our study and the reported Chinese CAD study. RESULTS Among the tRNA genes, we confirmed a highly conserved rare variant, A5592G, previously reported in the Chinese CAD study, and 2 novel rare mutations that reached Bonferroni's correction significance in the combined analysis were found (P=7.39×10-4 for T5628C in tRNAAla and P=1.01×10-5 for T681C in 12S rRNA) in the CAD study. Both of them were predicted to be pathological, with T5628C disrupting an extremely conservative base-pairing at the AC stem of tRNAAla. Furthermore, we confirmed the controversial issue that the number of non-synonymous heteroplasmic sites per sample was significantly higher in CAD patients. CONCLUSIONS In conclusion, our study confirmed the contribution of rare variants in CAD and showed that CAD patients had more non-synonymous heterogeneity mutations, which may be helpful in identifying the genetic and molecular basis of CAD.

Heart rhythm complexity as predictors for the prognosis of end-stage renal disease patients undergoing hemodialysis.

BACKGROUND: Heart rhythm complexity, a measure of heart rate dynamics and a risk predictor in various clinical diseases, has not been systematically studied in patients with end-stage renal disease. The aim of this study is to investigate the heart rhythm complexity and its prognostic value for mortality in end-stage renal disease patients undergoing hemodialysis. METHODS: To assess heart rhythm complexity and conventional heart rate variability measures, 4-h continuous electrocardiography for a retrospective cohort of 202 ostensibly healthy control subjects and 51 hemodialysis patients with end-stage renal disease were analyzed. Heart rhythm complexity was quantified by the complexity index from the measurement of the multiscale entropy profile. RESULTS: During a follow-up of 13 months, 8 people died in the patient group. Values of either traditional heart rate variability measurements or complexity indices were found significantly lower in patients than those in healthy controls. In addition, the complexity indices (Area 1-5, Area 6-15 and Area 6-20) in the mortality group were significantly lower than those in the survival group, while there were no significant differences in traditional heart rate variability parameters between the two groups. In receiver operating characteristic curve analysis, Area 6-20 (AUC = 0.895, p < 0.001) showed the strongest predictive power between mortality and survival groups. CONCLUSION: The results suggest that heart rhythm complexity is impaired for patients with end-stage renal disease. Furthermore, the complexity index of heart rate variability quantified by multiscale entropy may be a powerful independent predictor of mortality in end-stage renal disease patients undergoing hemodialysis.

HOXD-AS1/miR-130a sponge regulates glioma development by targeting E2F8.

Glioma development is an extremely complex process with changes occurring in numerous genes. HOXD antisense growth-associated long noncoding RNA (HOXD-AS1), an important long noncoding RNA (lncRNA), is known to regulate metastasis-related gene expression in bladder cancer, ovarian cancer and neuroblastoma. Here, we elucidated the function and possible molecular mechanisms of lncRNA HOXD-AS1 in human glioma cells. Our results proved that HOXD-AS1 expression was upregulated in glioma tissues and in glioma cell lines. HOXD-AS1 overexpression promoted cell migration and invasion in vitro, whereas knockdown of HOXD-AS1 expression repressed these cellular processes. Mechanistic studies further revealed that HOXD-AS1 could compete with the transcription factor E2F8 to bind with miR-130a, thus affecting E2F8 expression. Additionally, reciprocal repression was observed between HOXD-AS1 and miR-130a, and miR-130a mediated the tumor-suppressive effects of HOXD-AS1 knockdown. Taken together, these results provide a comprehensive analysis of the role of HOXD-AS1 in glioma cells and offer important clues to understand the key roles of competing endogenous RNA (ceRNA) mechanisms in human glioma.

Clinical and Biological Implications of Mutational Spectrum in Acute Myeloid Leukemia of FAB Subtypes M0 and M1.

BACKGROUND/AIMS: Acute myeloid leukemia (AML) of French-American-British (FAB) subtypes M0 and M1 are both poorly differentiated AML, but their mutational spectrum and molecular characteristics remain unknown. This study aimed to explore the mutational spectrum and prognostic factors of AML-M0 and M1. METHODS: Sixty-five AML patients derived from The Cancer Genome Atlas (TCGA) database were enrolled in this study. Whole-genome sequencing was performed to depict the mutational spectrum of each patient. Clinical characteristics at diagnosis, including peripheral blood (PB) white blood cell counts (WBC), blast percentages in PB and bone marrow (BM), FAB subtypes and the frequencies of known recurrent genetic mutations were described. Survival was estimated using the Kaplan-Meier methods and log-rank test. Univariate and multivariate Cox proportional hazard models were constructed for event-free survival (EFS) and overall survival (OS), using a limited backward elimination procedure. RESULTS: Forty-six patients had more than five recurrent genetic mutations. FLT3 had the highest mutation frequency (n=20, 31%), followed by NPM1 (n=18, 28%), DNMT3A (n=16, 25%), IDH1 (n=14, 22%), IDH2 (n=12, 18%), RUNX1 (n=11, 17%) and TET2 (n=7, 11%). Univariate analysis showed that age ≥60 years and TP53 mutations had adverse effect on EFS (P=0.015, P=0.036, respectively) and OS (P=0.003, P=0.004, respectively), WBC count ≥50×109/L and FLT3-ITD negatively affected EFS (P=0.003, P=0.034, respectively), whereas NPM1 mutations had favorable effect on OS (P=0.035) and allogeneic hematopoietic stem cell transplantation (allo-HSCT) on EFS and OS (all P< 0.001). Multivariate analysis suggested that allo-HSCT and NPM1 mutations were independent favorable prognostic factors for EFS and OS (all P< 0.05), WBC count ≥50×109/L was an independent risk factor for EFS (P=0.002) and TP53 mutations for OS (P=0.043). CONCLUSIONS: Our study provided new insights into the mutational spectrum and molecular signatures of AML-M0 and M1. We proposed that FLT3-ITD, NPM1 and TP53 be identified as markers for risk stratification of AML-M0 and M1. Patients with AML-M0 and M1 would likely benefit from allo-HSCT.

Combination Glioma Therapy Mediated by a Dual-Targeted Delivery System Constructed Using OMCN-PEG-Pep22/DOX.

Accumulating studies have investigated the efficacy of receptor-mediated delivery of hydrophobic drugs in glioma chemotherapy. Here, a delivery vehicle comprising polyethylene glycol (PEG) and oxidized nanocrystalline mesoporous carbon particles (OMCN) linked to the Pep22 polypeptide targeting the low-density lipoprotein receptor (LDLR) is designed to generate a novel drug-loaded system, designated as OMCN-PEG-Pep22/DOX (OPPD). This system effectively targets glioma cells and the blood-brain barrier and exerts therapeutic efficacy through both near-infrared (NIR) photothermal and chemotherapeutic effects of loaded doxycycline (DOX). Pathological tissue microarrays show an association of LDLR overexpression in human glioma tissue with patient survival.NIR irradiation treatment and magnetic resonance imaging results show that OPPD reaches the effective glioma-killing temperature in a glioma-bearing rat with a skull bone removal model and considerably reduces glioma sizes relative to the drug-loaded system without the Pep22 peptide modification and the control respectively. Thus, OPPD not only effectively targets LDLR-overexpressing glioma but also exerts a dual therapeutic effect by transporting DOX into the glioma and generating thermal effects with near-infrared irradiation to kill tumor cells. These collective findings support the utility of the novel OPPD drug-loaded system as a promising drug delivery vehicle for clinical application in glioma therapy.

BAALC and ERG expression levels at diagnosis have no prognosis impact on acute myeloid leukemia patients undergoing allogeneic hematopoietic stem cell transplantation.

Brain and acute leukemia, cytoplasmic (BAALC) and ETS-related gene (ERG) expression levels are independent prognostic factors for acute myeloid leukemia (AML); however, their prognostic impacts on AML patients undergoing allogeneic hematopoietic stem cell transplantation (allo-HSCT) require further investigation. We studied 71 de novo AML patients treated with allo-HSCT and defined low and high expressers according to the median expression levels of BAALC and ERG at diagnosis respectively. High BAALC expression was associated with wild-type NPM1 (P = 0.000) and RUNX1 mutations (P = 0.027). High ERG expression was associated with FLT3-ITD absence (P = 0.003) and wild-type NPM1 (P = 0.001). BAALC and ERG expression levels were significantly correlated with each other (P = 0.001). Survival analyses including Kaplan-Meier curves and univariate and multivariate analysis consistently reported that there were no significant differences for both event-free survival (EFS) and overall survival (OS) (all P > 0.1), between high versus low BAALC and ERG expressers. Our study suggested that despite of their well-known adverse role in prognosis of AML, neither BAALC nor ERG expression levels at diagnosis had effect on survival of AML patients who underwent allo-HSCT.