Search

PDF file - 589K, Supplementary Fig. S1 Elevated MDK levels in the culture medium of HCC cell lines compared with normal liver cell lines. Supplementary Fig. S2 The prognostic values of AFP and MDK levels for HCC patients. Supplementary Fig. S3.Expression of serum MDK in another independent cohort (cohort B) from Egypt. Supplementary Fig. S4. Logistic Regression Model to combine information of MDK and AFP for prediction of HCC in the learning cohort A (n=252): Comparison of AUROC for diagnosis HCC through logist regression model I, serum MDK and AFP. Supplementary Fig. S5. Positive predictive value (PPV) and negative predictive value (NPV) for identifying HCC patients through serum MDK according to different prevalence. Supplementary Fig. S6. Correlation between serum AFP and MDK levels in HCC patients (n=252). Supplementary Fig. S7. Serum MDK is elevated in different HCC subsets, No significant difference in serum MDK levels was found between the learning cohort A and the external validation cohort of HCC patients(P=0.398), or between the early stage HCC patients in learning and validation cohort (P=0.295). Table S1. Relationship between MDK expression and clinicopathological features of HCC patients. Table S2. Clinical and biochemical characteristics of liver cirrhosis (LC) patients in cohort A. Table S3. Baseline characteristics of the hospital controls with different etiologies from learning Cohort A. Table S4. Clinical characteristics of HCC patients according to different cohorts in this study. Table S5. MDK protein levels in cancer-free cirrhotic liver and HCC tissues. Table S6. Comparison of AFP and MDK with clinicopathological parameters in learning cohort A. Table S7．Univariate Cox-regression Analyses of Factors Associated With OS and TTR. Table S8. Sensitivities and specificities of AFP and MDK according to various cutoff values. Table S9. Cross-validation of the diagnostic ability of serum MDK. Table S10. Comparison of MDK with clinicopathological parameters in validation cohort C.

Purpose: To evaluate the value of serum midkine (MDK) as a diagnostic biomarker in hepatocellular carcinoma, particularly for those with negative alpha-fetoprotein (AFP) and at an early stage.Experimental Design: MDK expression in tumors was assessed by immunohistochemistry from 105 patients with hepatocellular carcinomas or liver cirrhosis. Serum MDK levels were detected by ELISA in 933 participants including hepatocellular carcinomas and hospital controls from different medical centers. Sensitivities and specificities of serum MDK in diagnosing hepatocellular carcinoma according to AFP level and Barcelona Clinic Liver Cancer (BCLC) stage were analyzed.Results: MDK levels were significantly elevated in hepatocellular carcinoma tissues as well as serum samples. The sensitivity of serum MDK for hepatocellular carcinoma diagnosis was much higher than that of AFP (86.9% vs. 51.9%) with similar specificities (83.9% vs. 86.3%). Notably, serum MDK had an outstanding performance in distinguishing AFP-negative hepatocellular carcinomas from different controls: In those AFP-negative hepatocellular carcinomas, the sensitivity could reach as high as 89.2%. Moreover, receiver operating characteristic (ROC) curve analysis also showed that serum MDK had a better performance compared with AFP in distinguishing early-stage hepatocellular carcinomas as well as small hepatocellular carcinomas. Even in very early-stage hepatocellular carcinomas, MDK showed an obviously higher sensitivity compared with AFP (80% vs. 40%). Furthermore, serum MDK level was significantly decreased in patients with hepatocellular carcinomas after curative resection and re-elevated when tumor relapse occurred.Conclusions: Serum MDK is significantly elevated in most hepatocellular carcinomas, including those with negative AFP and at an early stage, which may serve as a novel diagnostic marker in early diagnosis and postoperative monitoring of hepatocellular carcinomas. Clin Cancer Res; 19(14); 3944–54. ©2013 AACR.

Supplementary Tables 1-4, Figure Legend from Association of Specific Genotypes in Metastatic Suppressor HTPAP with Tumor Metastasis and Clinical Prognosis in Hepatocellular Carcinoma

The phosphatidic acid phosphatase HTPAP has been defined as a metastatic suppressor of hepatocellular carcinoma (HCC), but little is known about its function or potential applications as a prognostic marker. In this study, we analyzed patterns of HTPAP genetic variation and gene expression in 864 patients who underwent HCC resection, assessing these patterns for correlations to tumor metastasis potential. Focusing on two tagSNPs that were selected (+357G/C and +1838A/G), we found that only the +357G/C genotype was significantly associated with HTPAP mRNA and protein expression levels and the probability of metastasis. In an independent cohort of 665 HCC patients, we determined that the +357G/C genotype was associated with shorter time to recurrence and overall survival. Together, these results indicated that the HTPAP tagSNP +357 GG+GC genotypes may influence HCC metastatic potential and clinical prognosis by down-regulating HTPAP expression. Extending these results, a global expression profiling analysis identified 41 genes including the pro-inflammatory genes IL-8 and TLR2 that were significantly overexpressed in the +357 GG+GC group, as possible coregulated markers with HTPAP. Together, our findings identify an HTPAP genotype and associated gene expression pattern that favors metastasis progression and that could be used to predict tumor metastasis and prognosis in HCC patients. Cancer Res; 71(9); 3278–86. ©2011 AACR.

Supplementary Figure 1 from Association of Specific Genotypes in Metastatic Suppressor HTPAP with Tumor Metastasis and Clinical Prognosis in Hepatocellular Carcinoma

High-dimensional linear regression model is the most popular statistical model for high-dimensional data, but it is quite a challenging task to achieve a sparse set of regression coefficients. In this paper, we propose a simple heuristic algorithm to construct sparse high-dimensional linear regression models, which is adapted from the shortest solution-guided decimation algorithm and is referred to as ASSD. This algorithm constructs the support of regression coefficients under the guidance of the least-squares solution of the recursively decimated linear equations, and it applies an early-stopping criterion and a second-stage thresholding procedure to refine this support. Our extensive numerical results demonstrate that ASSD outperforms LASSO, vector approximate message passing, and two other representative greedy algorithms in solution accuracy and robustness. ASSD is especially suitable for linear regression problems with highly correlated measurement matrices encountered in real-world applications., 13 pages, 6 figures

Metastasis and chemo-resistance are still important factors that limit the overall efficacy of colorectal cancer treatment. Understanding the detailed molecular mechanism and identifying potential biomarkers are of great value in prognosis prediction and risk stratification. We investigated the role of miR-582-5p in colorectal cancer pathogenesis, progression and chemo-resistance. Furthermore, we explored the underlying molecular mechanism of miR-582-5p in modulation of malignant behaviors of colorectal cancer cells. Clinical samples and colorectal cancer cell lines were applied to explore miR-582-5p expression level and its significance on tumor cell metastasis and chemo-resistance. Transwell study and cellular survivability study were performed to explore the influences of miR-582-5p expression modulation on tumor cell chemo-resistance and invasion/migration. Dual-luciferase reporter gene assay was conducted to explore the influences of miR-582-5p on its target gene TNKS2. Colorectal cancer patients with lymph node or distal organ metastatic diseases exhibited significantly lower level of miR-582-5p. In vitro studies have indicated that miR-582-5p inhibition significantly increased migration and chemo-resistant capabilities of tumor cells. And dual-luciferase reporter gene assay demonstrated that miR-582-5p exhibited its influences on the biological behavior of tumor cells by targeting TNKS2. Our study demonstrated for the first time that miR-582-5p played an important role for colorectal tumor cell metastasis and chemo-resistance. Our research also indicated that miR-582-5p and its target gene TNKS2 could be novel biomarkers for metastatic disease prediction, overall prognosis evaluation, as well as potential therapeutic target for colorectal cancer patients.

To explore the role of transforming growth factor-β (TGF-β) in bladder neck contracture (BNC) after transurethral enucleation and resection of the prostate (TUERP).This study included 300 BPH patients undergoing TUERP, aged 51－89 (69.19 ± 8.43) years, with the prostate volume of 14.4－355.8 (63.18 ± 47.63) ml and preoperative IPSS of 15－35 (26.07 ± 5.9), QOL score of 3－6 (4.43 ± 0.67), PSA content of 0.17－23.16 (2.94 ± 3.77) ug/L, urinary leukocyte increase in 50 cases, post-void residual urine volume (PVR) of 0－440 (83.53 ± 86.85) ml, and maximum urinary flow rate (Qmax) of 2.3－14.5 (7.77 ± 3.47) ml/s. During TUERP, we collected the tissues from the bladder neck at 5 and 7 o'clock as well as the BPH tissue and the tissue from the residual prostate for HE staining, immunohistochemistry (the SP method) and examination of the infiltration degree of inflammatory cells and expressions of TGF-β1 and TGF-β3. During the 6－24 months follow-up, 6 of the patients were confirmed with BNC based on the clinical symptoms and the results of uroflowmetry and cystoscopy, and underwent transurethral bladder neck incision and detection of the expressions of TGF-β1 and TGF-β3 in the bladder neck tissue with BNC.The bladder neck tissue without BNC was mainly composed of smooth muscle and fibrous tissues with local infiltration of inflammatory cells, and the residual prostate tissue primarily comprised fibrous and muscle tissues, mixed with a little prostatic epithelial tissue. The bladder neck tissue with BNC, compared with that harvested during the initial TUERP, exhibited significantly increased expression of TGF-β1 (［68.20 ± 10.88］% vs ［36.14 ± 7.62］%, P0.05), decreased expression of TGF-β3 (［8.55 ± 4.73］% vs ［20.77 ± 8.69］%, P0.05), and enhanced infiltration of inflammatory cells (P0.05). The bladder neck tissue without BNC, in comparison with the BPH tissue, showed dramatically up-regulated expressions of TGF-β1 (［27.05 ± 8.21］% vs ［1.61 ± 0.69］%, P0.001) and TGF-β3 (［14.09 ± 4.19］% vs ［0.32 ± 0.11］%, P0.001) and increased infiltration of inflammatory cells (P0.05).After TUERP, the expression of TGF-β1 is increased, that of TGF-β3 decreased and the infiltration of inflammatory cells enhanced in the bladder neck tissue with BNC, which suggests that BNC may be related to the expression of TGF-β and that BNC after TUERP could be prevented by regulating the expression of TGF-β.

High-dimensional linear regression model is the most popular statistical model for high-dimensional data, but it is quite a challenging task to achieve a sparse set of regression coefficients. In this paper, we propose a simple heuristic algorithm to construct sparse high-dimensional linear regression models, which is adapted from the shortest-solution guided decimation algorithm and is referred to as ASSD. This algorithm constructs the support of regression coefficients under the guidance of the least-squares solution of the recursively decimated linear equations, and it applies an early-stopping criterion and a second-stage thresholding procedure to refine this support. Our extensive numerical results demonstrate that ASSD outper-forms LASSO, vector approximate message passing, and two other representative greedy algorithms in solution accuracy and robustness. ASSD is especially suitable for linear regression problems with highly correlated measurement matrices encountered in real-world applications.

Network dismantling aims at breaking a network into disconnected components and attacking vertices that intersect with many loops has proven to be a most efficient strategy. Yet existing loop-focusing methods do not distinguish the short loops within densely connected local clusters (e.g., cliques) from the long loops connecting different clusters, leading to lowered performance of these algorithms. Here we propose a new solution framework for network dismantling based on a two-scale bipartite factor-graph representation, in which long loops are maintained while local dense clusters are simplistically represented as individual factor nodes. A mean-field spin-glass theory is developed for the corresponding long-loop feedback vertex set problem. The framework allows for the advancement of various existing dismantling algorithms; we developed the new version of two benchmark algorithms BPD (which uses the message-passing equations of the spin-glass theory as the solver) and CoreHD (which is fastest among well-performing algorithms). New solvers outperform current state-of-the-art algorithms by a considerable margin on networks of various sorts. Further improvement in dismantling performance is achievable by opting flexibly the choice of local clusters.

We propose a method for solving statistical mechanics problems defined on sparse graphs. It extracts a small Feedback Vertex Set (FVS) from the sparse graph, converting the sparse system to a much smaller system with many-body and dense interactions with an effective energy on every configuration of the FVS, then learns a variational distribution parameterized using neural networks to approximate the original Boltzmann distribution. The method is able to estimate free energy, compute observables, and generate unbiased samples via direct sampling without auto-correlation. Extensive experiments show that our approach is more accurate than existing approaches for sparse spin glasses. On random graphs and real-world networks, our approach significantly outperforms the standard methods for sparse systems such as the belief-propagation algorithm; on structured sparse systems such as two-dimensional lattices our approach is significantly faster and more accurate than recently proposed variational autoregressive networks using convolution neural networks., 13 pages, 5 figures