A study of machine learning models for rapid intraoperative diagnosis of thyroid nodules for clinical practice in China.

BACKGROUND: In China, rapid intraoperative diagnosis of frozen sections of thyroid nodules is used to guide surgery. However, the lack of subspecialty pathologists and delayed diagnoses are challenges in clinical treatment. This study aimed to develop novel diagnostic approaches to increase diagnostic effectiveness. METHODS: Artificial intelligence and machine learning techniques were used to automatically diagnose histopathological slides. AI-based models were trained with annotations and selected as efficientnetV2-b0 from multi-set experiments. RESULTS: On 191 test slides, the proposed method predicted benign and malignant categories with a sensitivity of 72.65%, specificity of 100.0%, and AUC of 86.32%. For the subtype diagnosis, the best AUC was 99.46% for medullary thyroid cancer with an average of 237.6 s per slide. CONCLUSIONS: Within our testing dataset, the proposed method accurately diagnosed the thyroid nodules during surgery.

Whole human-brain mapping of single cortical neurons for profiling morphological diversity and stereotypy.

Quantifying neuron morphology and distribution at the whole-brain scale is essential to understand the structure and diversity of cell types. It is exceedingly challenging to reuse recent technologies of single-cell labeling and whole-brain imaging to study human brains. We propose adaptive cell tomography (ACTomography), a low-cost, high-throughput, and high-efficacy tomography approach, based on adaptive targeting of individual cells. We established a platform to inject dyes into cortical neurons in surgical tissues of 18 patients with brain tumors or other conditions and one donated fresh postmortem brain. We collected three-dimensional images of 1746 cortical neurons, of which 852 neurons were reconstructed to quantify local dendritic morphology, and mapped to standard atlases. In our data, human neurons are more diverse across brain regions than by subject age or gender. The strong stereotypy within cohorts of brain regions allows generating a statistical tensor field of neuron morphology to characterize anatomical modularity of a human brain.

Deep Learning-Based Recognition of Cervical Squamous Interepithelial Lesions.

Cervical squamous intraepithelial lesions (SILs) are precursor lesions of cervical cancer, and their accurate diagnosis enables patients to be treated before malignancy manifests. However, the identification of SILs is usually laborious and has low diagnostic consistency due to the high similarity of pathological SIL images. Although artificial intelligence (AI), especially deep learning algorithms, has drawn a lot of attention for its good performance in cervical cytology tasks, the use of AI for cervical histology is still in its early stages. The feature extraction, representation capabilities, and use of p16 immunohistochemistry (IHC) among existing models are inadequate. Therefore, in this study, we first designed a squamous epithelium segmentation algorithm and assigned the corresponding labels. Second, p16-positive area of IHC slides were extracted with Whole Image Net (WI-Net), followed by mapping the p16-positive area back to the H&E slides and generating a p16-positive mask for training. Finally, the p16-positive areas were inputted into Swin-B and ResNet-50 to classify the SILs. The dataset comprised 6171 patches from 111 patients; patches from 80% of the 90 patients were used for the training set. The accuracy of the Swin-B method for high-grade squamous intraepithelial lesion (HSIL) that we propose was 0.914 [0.889-0.928]. The ResNet-50 model for HSIL achieved an area under the receiver operating characteristic curve (AUC) of 0.935 [0.921-0.946] at the patch level, and the accuracy, sensitivity, and specificity were 0.845, 0.922, and 0.829, respectively. Therefore, our model can accurately identify HSIL, assisting the pathologist in solving actual diagnostic issues and even directing the follow-up treatment of patients.

STING agonism overcomes STAT3-mediated immunosuppression and adaptive resistance to PARP inhibition in ovarian cancer.

BACKGROUND: Poly (ADP-ribose) polymerase (PARP) inhibition (PARPi) has demonstrated potent therapeutic efficacy in patients with BRCA-mutant ovarian cancer. However, acquired resistance to PARPi remains a major challenge in the clinic. METHODS: PARPi-resistant ovarian cancer mouse models were generated by long-term treatment of olaparib in syngeneic Brca1-deficient ovarian tumors. Signal transducer and activator of transcription 3 (STAT3)-mediated immunosuppression was investigated in vitro by co-culture experiments and in vivo by analysis of immune cells in the tumor microenvironment (TME) of human and mouse PARPi-resistant tumors. Whole genome transcriptome analysis was performed to assess the antitumor immunomodulatory effect of STING (stimulator of interferon genes) agonists on myeloid cells in the TME of PARPi-resistant ovarian tumors. A STING agonist was used to overcome STAT3-mediated immunosuppression and acquired PARPi resistance in syngeneic and patient-derived xenografts models of ovarian cancer. RESULTS: In this study, we uncover an adaptive resistance mechanism to PARP inhibition mediated by tumor-associated macrophages (TAMs) in the TME. Markedly increased populations of protumor macrophages are found in BRCA-deficient ovarian tumors that rendered resistance to PARPi in both murine models and patients. Mechanistically, PARP inhibition elevates the STAT3 signaling pathway in tumor cells, which in turn promotes protumor polarization of TAMs. STAT3 ablation in tumor cells mitigates polarization of protumor macrophages and increases tumor-infiltrating T cells on PARP inhibition. These findings are corroborated in patient-derived, PARPi-resistant BRCA1-mutant ovarian tumors. Importantly, STING agonists reshape the immunosuppressive TME by reprogramming myeloid cells and overcome the TME-dependent adaptive resistance to PARPi in ovarian cancer. This effect is further enhanced by addition of the programmed cell death protein-1 blockade. CONCLUSIONS: We elucidate an adaptive immunosuppression mechanism rendering resistance to PARPi in BRCA1-mutant ovarian tumors. This is mediated by enrichment of protumor TAMs propelled by PARPi-induced STAT3 activation in tumor cells. We also provide a new strategy to reshape the immunosuppressive TME with STING agonists and overcome PARPi resistance in ovarian cancer.

KIF23 is a potential biomarker of diffuse large B cell lymphoma: Analysis based on bioinformatics and immunohistochemistry.

ABSTRACT: Diffuse Large B Cell Lymphoma (DLBCL), the most common form of blood cancer. The genetic and clinical heterogeneity of DLBCL poses a major barrier to diagnosis and treatment. Hence, we aim to identify potential biomarkers for DLBCL.Differentially expressed genes were screened between DLBCL and the corresponding normal tissues. Kyoto Encyclopedia of Genes and Genomes and Gene oncology analyses were performed to obtain an insight into these differentially expressed genes. PPI network was constructed to identify hub genes. survival analysis was applied to evaluate the prognostic value of those hub genes. DNA methylation analysis was implemented to explore the epigenetic dysregulation of genes in DLBCL.In this study, Kinesin family member 23 (KIF23) showed higher expression in DLBCL and was identified as a risk factor in DLBCL. The immunohistochemistry experiment further confirmed this finding. Subsequently, the univariate and multivariate analysis indicated that KIF23 might be an independent adverse factor in DLBCL. Upregulation of KIF23 might be a risk factor for the overall survival of patients who received an R-CHOP regimen, in late-stage, whatever with or without extranodal sites. Higher expression of KIF23 also significantly reduced 3, 5, 10-year overall survival. Furthermore, functional enrichment analyses (Kyoto Encyclopedia of Genes and Genomes, Gene oncology, and Gene Set Enrichment Analysis) showed that KIF23 was mainly involved in cell cycle, nuclear division, PI3K/AKT/mTOR, TGF-beta, and Wnt/beta-catenin pathway in DLBCL. Finally, results of DNA methylation analysis indicated that hypomethylation in KIF23's promoter region might be the result of its higher expression in DLBCL.The findings of this study suggested that KIF23 is a potential biomarker for the diagnosis and prognosis of DLBCL. However, further studies were needed to validate these findings.

STING agonism reprograms tumor-associated macrophages and overcomes resistance to PARP inhibition in BRCA1-deficient models of breast cancer.

PARP inhibitors (PARPi) have drastically changed the treatment landscape of advanced ovarian tumors with BRCA mutations. However, the impact of this class of inhibitors in patients with advanced BRCA-mutant breast cancer is relatively modest. Using a syngeneic genetically-engineered mouse model of breast tumor driven by Brca1 deficiency, we show that tumor-associated macrophages (TAMs) blunt PARPi efficacy both in vivo and in vitro. Mechanistically, BRCA1-deficient breast tumor cells induce pro-tumor polarization of TAMs, which in turn suppress PARPi-elicited DNA damage in tumor cells, leading to reduced production of dsDNA fragments and synthetic lethality, hence impairing STING-dependent anti-tumor immunity. STING agonists reprogram M2-like pro-tumor macrophages into an M1-like anti-tumor state in a macrophage STING-dependent manner. Systemic administration of a STING agonist breaches multiple layers of tumor cell-mediated suppression of immune cells, and synergizes with PARPi to suppress tumor growth. The therapeutic benefits of this combination require host STING and are mediated by a type I IFN response and CD8+ T cells, but do not rely on tumor cell-intrinsic STING. Our data illustrate the importance of targeting innate immune suppression to facilitate PARPi-mediated engagement of anti-tumor immunity in breast cancer.

NEK2 Promotes Cell Proliferation and Glycolysis by Regulating PKM2 Abundance via Phosphorylation in Diffuse Large B-Cell Lymphoma.

Diffuse large B-cell lymphoma (DLBCL) is the most frequent and commonly diagnosed subtype of NHL, which is characterized by high heterogeneity and malignancy, and most DLBCL patients are at advanced stages. The serine/threonine kinase NEK2 (NIMA-related kinase 2), a member of NIMA-related kinase (NEK) family that regulates cell cycle, is upregulated in a variety of malignancies, including diffuse large B-cell lymphoma. However, the role and underlying mechanisms of NEK2 in DLBCL have seldom been discussed. In this study, we identified that NEK2 is upregulated in DLBCL compared to normal lymphoid tissues, and overexpression of NEK2 predicted a worse prognosis of DLBCL patients. Gene set enrichment analysis indicates that NEK2 might participate in regulating glycolysis. Knockdown of NEK2 inhibited growth and glycolysis of DLBCL cells. The interaction between NEK2 and PKM2 was discovered by tandem affinity purification and then was confirmed by immunofluorescence staining, coimmunoprecipitation, and immunoprecipitation. NEK2 bounds to PKM2 and regulates PKM2 abundance via phosphorylation, which increases PKM2 stability. The xenograft tumor model checks the influence of NEK2 on tumor growth in vivo. Thus, NEK2 could be the novel biomarker and target of DLBCL, which remarkably ameliorates the diagnosis and treatment of DLBCL.

Identification of hub genes associated with the pathogenesis of diffuse large B-cell lymphoma subtype one characterized by host response via integrated bioinformatic analyses.

BACKGROUND: Host response diffuse large B-cell lymphoma (HR DLBCL) shares features of histologically defined T-cell/histiocyte-rich B-cell lymphoma, including fewer genetic abnormalities, frequent splenic and bone marrow involvement, and younger age at presentation. HR DLBCL is inherently less responsive to the standard treatment for DLBCL. Moreover, the mechanism of infiltration of HR DLBCL with preexisting abundant T-cells and dendritic cells is unknown, and their associated underlying immune responses incompletely defined. Here, hub genes and pathogenesis associated with HR DLBCL were explored to reveal molecular mechanisms and treatment targets. METHODS: Differentially expressed genes were identified in three datasets (GSE25638, GSE44337, GSE56315). The expression profile of the genes in the GSE53786 dataset was used to constructed a co-expression network. Protein-protein interactions analysis in the modules of interest identified candidate hub genes. Then screening of real hub genes was carried out by survival analysis within the GSE53786 and GSE10846 datasets. Expression of hub genes was validated in the Gene expression profiling interactive analysis, Oncomine databases and human tissue specimens. Functional enrichment analysis and Gene set enrichment analysis were utilized to investigate the potential mechanisms. Tumor Immune Estimation Resource and The Cancer Genome Atlas were used to mine the association of the hub gene with tumor immunity, potential upstream regulators were predicted using bioinformatics tools. RESULTS: A total of 274 common differentially expressed genes were identified. Within the key module, we identified CXCL10 as a real hub gene. The validation of upregulated expression level of CXCL10 was consistent with our study. CXCL10 might have a regulatory effect on tumor immunity. The predicted miRNA (hsa-mir-6849-3p) and transcription factor (IRF9) might regulate gene expression in the hub module.

EZH2 cooperates with gain-of-function p53 mutants to promote cancer growth and metastasis.

In light of the increasing number of identified cancer-driven gain-of-function (GOF) mutants of p53, it is important to define a common mechanism to systematically target several mutants, rather than developing strategies tailored to inhibit each mutant individually. Here, using RNA immunoprecipitation-sequencing (RIP-seq), we identified the Polycomb-group histone methyltransferase EZH2 as a p53 mRNA-binding protein. EZH2 bound to an internal ribosome entry site (IRES) in the 5'UTR of p53 mRNA and enhanced p53 protein translation in a methyltransferase-independent manner. EZH2 augmented p53 GOF mutant-mediated cancer growth and metastasis by increasing protein levels of mutant p53. EZH2 overexpression was associated with worsened outcome selectively in patients with p53-mutated cancer. Depletion of EZH2 by antisense oligonucleotides inhibited p53 GOF mutant-mediated cancer growth. Our findings reveal a non-methyltransferase function of EZH2 that controls protein translation of p53 GOF mutants, inhibition of which causes synthetic lethality in cancer cells expressing p53 GOF mutants.

PARP Inhibition Elicits STING-Dependent Antitumor Immunity in Brca1-Deficient Ovarian Cancer.

PARP inhibitors have shown promising clinical activities for patients with BRCA mutations and are changing the landscape of ovarian cancer treatment. However, the therapeutic mechanisms of action for PARP inhibition in the interaction of tumors with the tumor microenvironment and the host immune system remain unclear. We find that PARP inhibition by olaparib triggers robust local and systemic antitumor immunity involving both adaptive and innate immune responses through a STING-dependent antitumor immune response in mice bearing Brca1-deficient ovarian tumors. This effect is further augmented when olaparib is combined with PD-1 blockade. Our findings thus provide a molecular mechanism underlying antitumor activity by PARP inhibition and lay a foundation to improve therapeutic outcome for cancer patients.

Wilms' tumor 1-associating protein plays an aggressive role in diffuse large B-cell lymphoma and forms a complex with BCL6 via Hsp90.

BACKGROUND: Wilms' tumor 1-associating protein (WTAP) is a nuclear protein, which is ubiquitously expressed in many tissues. Furthermore, in various types of malignancies WTAP is overexpressed and plays a role as an oncogene. The function of WTAP in diffuse large B-cell lymphoma (DLBCL), however, remains unclear. METHODS: Immunohistochemistry was applied to evaluate the levels of WTAP expression in DLBCL tissues and normal lymphoid tissues. Overexpression and knock-down of WTAP in DLBCL cell lines, verified on mRNA and protein level served to analyze cell proliferation and apoptosis in DLBCL cell lines by flow cytometry. Finally, co-immunoprecipitation (Co-IP), IP, and GST-pull down assessed the interaction of WTAP with Heat shock protein 90 (Hsp90) and B-cell lymphoma 6 (BCL6) as well as determined the extend of its ubiquitinylation. RESULTS: WTAP protein levels were consistently upregulated in DLBCL tissues. WTAP promoted DLBCL cell proliferation and improved the ability to confront apoptosis, while knockdown of WTAP in DLBCL cell lines allowed a significant higher apoptosis rate after treatment with Etoposide, an anti-tumor drug. The stable expression of WTAP was depended on Hsp90. In line, we demonstrated that WTAP could form a complex with BCL6 via Hsp90 in vivo and in vitro. CONCLUSION: WTAP is highly expressed in DLBCL, promoting growth and anti-apoptosis in DLBCL cell lines. WTAP is a client protein of Hsp90 and can appear in a complex with BCL6 and Hsp90 in DLBCL. Down-regulation of WTAP could improve the chemotherapeutic treatments in DLBCL.

Inhibition of EZH2 by chemo- and radiotherapy agents and small molecule inhibitors induces cell death in castration-resistant prostate cancer.

Androgen deprivation therapy is the mainstay of treatment of advanced prostate cancer (PCa). However, a significant portion of patients experience disease relapse and tumors ultimately evolve into castration resistant prostate cancer (CRPC), for which there is no cure in the clinic. The Polycomb protein enhancer of zeste homolog 2 (EZH2) is frequently overexpressed in CRPC. It is unclear whether EZH2 can be a therapeutic target in CRPC. Here, we demonstrated that chemo- and radiotherapy agents such as camptothecin (CPT) and γ irradiation decrease EZH2 expression in various PCa cell lines. We provided evidence that functional p53 and RB proteins are required for CPT- and irradiation-induced downregulation of EZH2 in CRPC cells. We demonstrated that EZH2-specific small molecule inhibitors mitigate CRPC cell growth. We further showed that the EZH2 inhibitor GSK126 inhibits both Polycomb-dependent and -independent functions of EZH2 in PCa cells. Importantly, we found that inhibition of EZH2 by genetic and pharmacological means sensitizes CRPC cells to CPT-induced apoptotic death and growth inhibition in culture and in mice. Our data suggest that concomitant administration of small molecule inhibitors of EZH2 may significantly increase the anti-tumor efficacy of conventional chemo- and radiotherapies in CRPC.

LncRNA MALAT1 enhances oncogenic activities of EZH2 in castration-resistant prostate cancer.

The Polycomb protein enhancer of zeste homolog 2 (EZH2) is frequently overexpressed in advanced human prostate cancer (PCa), especially in lethal castration-resistant prostate cancer (CRPC). However, the signaling pathways that regulate EZH2 functions in PCa remain incompletely defined. Using EZH2 antibody-based RNA immunoprecipitation-coupled high throughput sequencing (RIP-seq), we demonstrated that EZH2 binds to MALAT1, a long non-coding RNA (lncRNA) that is overexpressed during PCa progression. GST pull-down and RIP assays demonstrated that the 3' end of MALAT1 interacts with the N-terminal of EZH2. Knockdown of MALAT1 impaired EZH2 recruitment to its target loci and upregulated expression of EZH2 repressed genes. Further studies indicated that MALAT1 plays a vital role in EZH2-enhanced migration and invasion in CRPC cell lines. Meta-analysis and RT-qPCR of patient specimens demonstrated a positive correlation between MALAT1 and EZH2 expression in human CRPC tissues. Finally, we showed that MALAT1 enhances expression of PRC2-independent target genes of EZH2 in CRPC cells in culture and patient-derived xenografts. Together, these data indicate that MALAT1 may be a crucial RNA cofactor of EZH2 and that the EZH2-MALAT1 association may provide a new avenue for development new strategies for treatment of CRPC.

Suppression of casein kinase 2 sensitizes tumor cells to antitumor TRAIL therapy by regulating the phosphorylation and localization of p65 in prostate cancer.

In the United States, prostate cancer (PCa) is the most commonly diagnosed cancer in males. For PCa at the late hormone-refractory stage, substantial improvement in treatment strategies is critically needed. TNF-related apoptosis-inducing ligand (TRAIL) is a promising anticancer agent, but both intrinsic and acquired resistance to TRAIL poses a huge problem in establishing clinically effective TRAIL therapies. In the present study, we examined the role played by casein kinase 2 (CK2) in the TRAIL­induced nuclear factor κ-light-chain-enhancer of activated B cell (NF-κB) pathway in a PCa cell line. Downregulation of CK2 combined with a sub-dose of TRAIL suppressed p65 phosphorylation at serine 536. The combination treatment of TRAIL and the CK2 inhibitor decreased p65 nuclear translocation. Under the treatment of a sub-dose of TRAIL, downregulation of CK2, using both genetic and pharmacological approaches, decreased the transcriptional activity of NF-κB and the expression of NF-κB downstream anti-apoptosis genes. Therefore, we provided novel molecular mechanistic insight reporting that CK2 regulates the sensitivity of PCa cells to the antitumor effect of TRAIL. This is important for understanding how the TRAIL pathway is disrupted in PCa and may help to develop an effective combinatorial therapy for PCa.

p300 acetyltransferase regulates androgen receptor degradation and PTEN-deficient prostate tumorigenesis.

Overexpression of the histone acetyltransferase p300 is implicated in the proliferation and progression of prostate cancer, but evidence of a causal role is lacking. In this study, we provide genetic evidence that this generic transcriptional coactivator functions as a positive modifier of prostate tumorigenesis. In a mouse model of PTEN deletion-induced prostate cancer, genetic ablation of p300 attenuated expression of the androgen receptor (AR). This finding was confirmed in human prostate cancer cells in which PTEN expression was abolished by RNA interference-mediated attenuation. These results were consistent with clinical evidence that the expression of p300 and AR correlates positively in human prostate cancer specimens. Mechanistically, PTEN inactivation increased AR phosphorylation at serine 81 (Ser81) to promote p300 binding and acetylation of AR, thereby precluding its polyubiquitination and degradation. In support of these findings, in PTEN-deficient prostate cancer in the mouse, we found that p300 was crucial for AR target gene expression. Taken together, our work identifies p300 as a molecular determinant of AR degradation and highlights p300 as a candidate target to manage prostate cancer, especially in cases marked by PTEN loss.

CBP loss cooperates with PTEN haploinsufficiency to drive prostate cancer: implications for epigenetic therapy.

Despite the high incidence and mortality of prostate cancer, the etiology of this disease is not fully understood. In this study, we develop functional evidence for CBP and PTEN interaction in prostate cancer based on findings of their correlate expression in the human disease. Cbp(pc-/-);Pten(pc+/-) mice exhibited higher cell proliferation in the prostate and an early onset of high-grade prostatic intraepithelial neoplasia. Levels of EZH2 methyltransferase were increased along with its Thr350 phosphorylation in both mouse Cbp(-/-); Pten(+/-) and human prostate cancer cells. CBP loss and PTEN deficiency cooperated to trigger a switch from K27-acetylated histone H3 to K27-trimethylated bulk histones in a manner associated with decreased expression of the growth inhibitory EZH2 target genes DAB2IP, p27(KIP1), and p21(CIP1). Conversely, treatment with the histone deacetylase inhibitor panobinostat reversed this switch, in a manner associated with tumor suppression in Cbp(pc-/-);Pten(pc+/-) mice. Our findings show how CBP and PTEN interact to mediate tumor suppression in the prostate, establishing a central role for histone modification in the etiology of prostate cancer and providing a rationale for clinical evaluation of epigenetic-targeted therapy in patients with prostate cancer.

BRCA1 is a negative modulator of the PRC2 complex.

The Polycomb-repressive complex 2 (PRC2) is important for maintenance of stem cell pluripotency and suppression of cell differentiation by promoting histone H3 lysine 27 trimethylation (H3K27me3) and transcriptional repression of differentiation genes. Here we show that the tumour-suppressor protein BRCA1 interacts with the Polycomb protein EZH2 in mouse embryonic stem (ES) and human breast cancer cells. The BRCA1-binding region in EZH2 overlaps with the noncoding RNA (ncRNA)-binding domain, and BRCA1 expression inhibits the binding of EZH2 to the HOTAIR ncRNA. Decreased expression of BRCA1 causes genome-wide EZH2 re-targeting and elevates H3K27me3 levels at PRC2 target loci in both mouse ES and human breast cancer cells. BRCA1 deficiency blocks ES cell differentiation and enhances breast cancer migration and invasion in an EZH2-dependent manner. These results reveal that BRCA1 is a key negative modulator of PRC2 and that loss of BRCA1 inhibits ES cell differentiation and enhances an aggressive breast cancer phenotype by affecting PRC2 function.

Using gene co-expression network analysis to predict biomarkers for chronic lymphocytic leukemia.

BACKGROUND: Chronic lymphocytic leukemia (CLL) is the most common adult leukemia. It is a highly heterogeneous disease, and can be divided roughly into indolent and progressive stages based on classic clinical markers. Immunoglobin heavy chain variable region (IgVH) mutational status was found to be associated with patient survival outcome, and biomarkers linked to the IgVH status has been a focus in the CLL prognosis research field. However, biomarkers highly correlated with IgVH mutational status which can accurately predict the survival outcome are yet to be discovered. RESULTS: In this paper, we investigate the use of gene co-expression network analysis to identify potential biomarkers for CLL. Specifically we focused on the co-expression network involving ZAP70, a well characterized biomarker for CLL. We selected 23 microarray datasets corresponding to multiple types of cancer from the Gene Expression Omnibus (GEO) and used the frequent network mining algorithm CODENSE to identify highly connected gene co-expression networks spanning the entire genome, then evaluated the genes in the co-expression network in which ZAP70 is involved. We then applied a set of feature selection methods to further select genes which are capable of predicting IgVH mutation status from the ZAP70 co-expression network. CONCLUSIONS: We have identified a set of genes that are potential CLL prognostic biomarkers IL2RB, CD8A, CD247, LAG3 and KLRK1, which can predict CLL patient IgVH mutational status with high accuracies. Their prognostic capabilities were cross-validated by applying these biomarker candidates to classify patients into different outcome groups using a CLL microarray datasets with clinical information.

Features versus context: An approach for precise and detailed detection and delineation of faces and facial features.

The appearance-based approach to face detection has seen great advances in the last several years. In this approach, we learn the image statistics describing the texture pattern (appearance) of the object class we want to detect, e.g., the face. However, this approach has had limited success in providing an accurate and detailed description of the internal facial features, i.e., eyes, brows, nose, and mouth. In general, this is due to the limited information carried by the learned statistical model. While the face template is relatively rich in texture, facial features (e.g., eyes, nose, and mouth) do not carry enough discriminative information to tell them apart from all possible background images. We resolve this problem by adding the context information of each facial feature in the design of the statistical model. In the proposed approach, the context information defines the image statistics most correlated with the surroundings of each facial component. This means that when we search for a face or facial feature, we look for those locations which most resemble the feature yet are most dissimilar to its context. This dissimilarity with the context features forces the detector to gravitate toward an accurate estimate of the position of the facial feature. Learning to discriminate between feature and context templates is difficult, however, because the context and the texture of the facial features vary widely under changing expression, pose, and illumination, and may even resemble one another. We address this problem with the use of subclass divisions. We derive two algorithms to automatically divide the training samples of each facial feature into a set of subclasses, each representing a distinct construction of the same facial component (e.g., closed versus open eyes) or its context (e.g., different hairstyles). The first algorithm is based on a discriminant analysis formulation. The second algorithm is an extension of the AdaBoost approach. We provide extensive experimental results using still images and video sequences for a total of 3,930 images. We show that the results are almost as good as those obtained with manual detection.

Carbonyl reductase 1 as a novel target of (-)-epigallocatechin gallate against hepatocellular carcinoma.

UNLABELLED: Human carbonyl reductase 1 (CBR1) converts the antitumor drug and anthracycline daunorubicin (DNR) into the alcohol metabolite daunorubicinol (DNROL) with significantly reduced antitumor activity and cardiotoxicity, and this limits the clinical use of DNR. Inhibition of CBR1 can thus increase the efficacy and decrease the toxicity of DNR. Here we report that (-)-epigallocatechin gallate (EGCG) from green tea is a promising inhibitor of CBR1. EGCG directly interacts with CBR1 and acts as a noncompetitive inhibitor with respect to the cofactor reduced nicotinamide adenine dinucleotide phosphate and the substrate isatin. The inhibition is dependent on the pH, and the gallate moiety of EGCG is required for activity. Molecular modeling has revealed that EGCG occupies the active site of CBR1. Furthermore, EGCG specifically enhanced the antitumor activity of DNR against hepatocellular carcinoma SMMC7721 cells expressing high levels of CBR1 and corresponding xenografts. We also demonstrated that EGCG could overcome the resistance to DNR by Hep3B cells stably expressing CBR1 but not by RNA interference of CBR1-HepG2 cells. The level of the metabolite DNROL was negatively correlated with that of EGCG in the cell extracts. Finally, EGCG decreased the cardiotoxicity of DNR in a human carcinoma xenograft model with both SMMC7721 and Hep3B cells in mice. CONCLUSION: These results strongly suggest that EGCG can inhibit CBR1 activity and enhance the effectiveness and decrease the cardiotoxicity of the anticancer drug DNR. These findings also indicate that a combination of EGCG and DNR might represent a novel approach for hepatocellular carcinoma therapy or chemoprevention.