Biology subjects

Hwang, T. H.

Publications and source records attributed to Hwang, T. H..

4 recordsLinked to original sources

bioRxiv2018

Divine: Prioritizing Genes for Rare Mendelian Disease in Whole Exome Sequencing Data

MotivationRecent studies showed that a phenotype-driven analysis of whole exome sequencing (WES) could provide more accurate and clinically relevant genetic variants.\n\nResultsWe develop a computational tool called Divine that integrates patients phenotype(s) and WES data with 30 prior biological knowledge (e.g., human phenotype ontology, gene ontology, pathway database, protein-protein interaction networks, pathogenicity by the amino acid change due to polymorphism, and hot-spot protein domains) to prioritize potential disease-causing genes. In a retrospective study with 22 real and four simulated data set, Divine ranks the same pathogenic genes confirmed by the original studies 5th on average out of a thousand of mutated genes and outperforms existing state-of-the-art methods.\n\nAvailabilityhttps://github.com/hwanglab/divine\n\nContacthwangt@ccf.org\n\nSupplementary informationSupplementary Document is attached at the end of the page.

bioinformatics↗

bioRxiv2018

Automatic Classification of Prostate Cancer Gleason Scores from Digitized Whole Slide Tissue Biopsies

Histological Gleason grading of tumor patterns is one of the most powerful prognostic predictors in prostate cancer. However, manual analysis and grading performed by pathologists are typically subjective and time-consuming. In this paper, we propose an automatic technique for Gleason grading of prostate cancer from H&E stained whole slide biopsy images using a set of novel completed and statistical local bi-nary pattern (CSLBP) descriptors. First the technique divides the whole slide image into a set of small image tiles, where salient tumor tiles with high nuclei densities are selected for analysis. The CSLBP texture features that encode pixel intensity variations from circularly surrounding neighborhoods are then extracted from salient image tiles to characterize different Gleason patterns. Finally, CSLBP texture features computed from all tiles are integrated and utilized by the multi-class support vector machine (SVM) that assigns patient biopsy with different Gleason score of 6, 7 or [≥]8. Experiments have been performed on 312 different patient cases selected from the cancer genome atlas (TCGA) and have achieved more than 79% classification accuracies, which is superior to state-of-the-art textural descriptors for prostate cancer Gleason grading.

bioinformatics↗

bioRxiv2018

Molecular Subtypes of Anaplastic Gliomas Identified with Somatic-Mutation and Pathway Based Gene Signature

PurposeAnaplastic gliomas constitute heterogeneous population with variable outcomes and no consensus on therapeutic approach. Molecular profiling may provide prognostication beyond clinical and pathologic factors and help guide treatment decisions.\n\nExperimental DesignThe Cancer Genome Atlas (TCGA) was utilized to derive a 39-gene low grade glioma-specific gene signature. Consensus clustering based on expression of the signature identified subgroups for 176 patients with anaplastic glioma from TCGA. Overall survival (OS) was analyzed for each subgroup. A total of 68 patients from Repository for Molecular Brain Neoplasia Data (REMBRANDT) were used as an independent validation dataset.\n\nResultsConsensus clustering separated the TCGA group into two distinct cohorts. The OS was significantly different between two subgroups, 20 vs. 67 months (p<0.001). On univariate analysis, the molecular subgroup, age, KPS, IDH1/2 mutation, 1p19q-co-deletion, chemotherapy, and use of both chemotherapy and radiation-therapy were significantly associated with OS. On multivariable analysis, the molecular subgroup remained significant with HR of 2.6 (p=.047, 95%CI [1.01-6.68]). In an independent validation with REMBRANDT, consensus clustering based on the signature successfully identified similarly poor prognostic subgroup with median survival of 14 months and concordance of expression patterns in 21 of the genes.\n\nConclusionExpression patterns of the 39 gene stratified anaplastic gliomas into two distinct subgroups with substantially different OS. This molecular prognostication was validated in an external dataset. Utilization of molecular subgroup, in addition to known prognostic factors may help define those requiring aggressive therapeutic intervention. Characteristic genes within the poor prognostic group may represent potential targets for therapeutic intensification.\n\nSource code and dataset used in this work is available for reviewers at: https://www.taehyunlab.org/ntripath\n\nImportance of the studyDespite the revolution of tailored therapy, anaplastic gliomas represent a category of tumors without clear treatment recommendations. While current prognostic factors help guide therapy recommendations, further refinement with the addition of molecular markers can help physicians with treatment recommendations. In this study, we developed a 39 gene prognostic gene signature by utilizing Pan-Cancer TCGA mutation profiles of over 5,000 patients across 19 different TCGA cancer types to stratify patients with anaplastic gliomas. We performed consensus clustering based on gene expression profiles of our 39-gene signature without any consideration of clinical factors or outcomes to TCGA anaplastic gliomas patients as well as an independent dataset and successfully identified two molecular subgroups with distinct clinical outcome. We found that subgroups have remarkably different survivals with a clear poor prognostic group. Furthermore, the poor prognostic group showed significant benefits for aggressive multimodality therapy, justifying the use of intensive therapy based on molecular stratification.

genomics↗

bioRxiv2017

miR-16 suppresses growth of rhabdoid tumor cells

BackgroundRhabdoid tumor is a highly aggressive pediatric cancer characterized by biallelic loss and/or mutation of SMARCB1. Outcomes remain poor, and there are no established ways to target the tumorigenic pathways driven by SMARCB1 inactivation. SMARCB1 loss leads to an increase in cyclin D transcription.\n\nProcedureWe characterized the cell line WT-CLS1, which has been described previously as Wilms tumor, by whole-exome sequencing, RNA-seq, and xenograft histology. We measured the effect of microRNA overexpression on WT-CLS1, BT-12, and CHLA-06-ATRT.\n\nResultsWe found that WT-CLS1 demonstrates the histological, mutational, and transcriptional hallmarks of rhabdoid tumor. Because the microRNAs let-7 and miR-16 can target cyclin D genes, we next overexpressed each of these microRNAs in WT-CLS1. We found that miR-16 reduced cell accumulation. This was accompanied by a decrease in proliferation markers and an increase in apoptosis markers. These results were replicated in the BT-12 and CHLA-06-ATRT cell lines.\n\nConclusionsThe loss-of-function SMARCB1 mutation found in WT-CLS1, in conjunction with immunohistochemical and gene expression analysis, warrants reclassification of this cell line as rhabdoid tumor. Proliferation of WT-CLS1 and other rhabdoid tumor cell lines is significantly abrogated by miR-16 overexpression. Further studies are necessary to gain insight into the potential for miR-16 to be used as a novel therapeutic in rhabdoid tumor.\n\nAbbreviations

cancer biology↗