Biclustering Analysis for Pattern Discovery: Current Techniques, Comparative Studies and Applications

Zhao, Hongya; Liew, Alan Wee‐Chung; Wang, Doris Z.; Yan, Hong

doi:10.2174/157489312799304413

Cited by 48 publications

(27 citation statements)

References 82 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Thus, biclustering has become a popular technique and lots of algorithms are proposed, such as distance-based [ 28 , 29 ], factorization-based [ 30 , 31 ] and geometric-based biclustering [ 32 , 33 ]. Most biclustering algorithms [ 34 – 38 ] allow bi-clusters to have partially overlap, and some objects (samples or genes) may not belong to any bi-cluster at all [ 39 , 40 ]. This character of biclustering, although useful in some instances [ 26 ], is not good for interpretation.…”

Section: Introductionmentioning

confidence: 99%

Gene expression profiling of 1200 pancreatic ductal adenocarcinoma reveals novel subtypes

Zhao

Yan

2018

BMC Cancer

Self Cite

View full text Add to dashboard Cite

BackgroundPancreatic ductal adenocarcinoma (PDAC) is the fourth leading cause of cancer related death in the world with a five-year survival rate of less than 5%. Not all PDAC are the same, because there exist intra-tumoral heterogeneity between PDAC, which poses a great challenge to personalized treatments for PDAC.MethodsTo dissect the molecular heterogeneity of PDAC, we performed a retrospective meta-analysis on whole transcriptome data from more than 1200 PDAC patients. Subtypes were identified based on non-negative matrix factorization (NMF) biclustering method. We used the gene set enrichment analysis (GSEA) and survival analysis to conduct the molecular and clinical characterization of the identified subtypes, respectively.ResultsSix molecular and clinical distinct subtypes of PDAC: L1-L6, are identified and grouped into tumor-specific (L1, L2 and L6) and stroma-specific subtypes (L3, L4 and L5). For tumor-specific subtypes, L1 (~ 22%) has enriched carbohydrate metabolism-related gene sets and has intermediate survival. L2 (~ 22%) has the worst clinical outcomes, and is enriched for cell proliferation-related gene sets. About 23% patients can be classified into L6, which leads to intermediate survival and is enriched for lipid and protein metabolism-related gene sets. Stroma-specific subtypes may contain high non-epithelial contents such as collagen, immune and islet cells, respectively. For instance, L3 (~ 12%) has poor survival and is enriched for collagen-associated gene sets. L4 (~ 14%) is enriched for various immune-related gene sets and has relatively good survival. And L5 (~ 7%) has good clinical outcomes and is enriched for neurotransmitter and insulin secretion related gene sets. In the meantime, we identified 160 subtype-specific markers and built a deep learning-based classifier for PDAC. We also applied our classification system on validation datasets and observed much similar molecular and clinical characteristics between subtypes.ConclusionsOur study is the largest cohort of PDAC gene expression profiles investigated so far, which greatly increased the statistical power and provided more robust results. We identified six molecular and clinical distinct subtypes to describe a more complete picture of the PDAC heterogeneity. The 160 subtype-specific markers and a deep learning based classification system may be used to better stratify PDAC patients for personalized treatments.Electronic supplementary materialThe online version of this article (10.1186/s12885-018-4546-8) contains supplementary material, which is available to authorized users.

show abstract

Section: Introductionmentioning

confidence: 99%

Gene expression profiling of 1200 pancreatic ductal adenocarcinoma reveals novel subtypes

Zhao

Yan

2018

BMC Cancer

Self Cite

View full text Add to dashboard Cite

show abstract

“…perhaps psychotropic medication and referral to specialist are often used separately and seldom used together), as well as practise patterns that are seldom or never used. We will also use a bi-clustering technique (sometimes called co-clustering or block clustering) 40 to simultaneously identify subgroups of family physicians who have similar practise patterns based on available data, and subgroups of patients more likely to be managed in primary care.…”

Section: Methodsmentioning

confidence: 99%

Understanding the role of the family physician in early psychosis intervention

et al. 2018

View full text Add to dashboard Cite

BackgroundThe family physician is key to facilitating access to psychiatric treatment for young people with first-episode psychosis, and this involvement can reduce aversive events in pathways to care. Those who seek help from primary care tend to have longer intervals to psychiatric care, and some people receive ongoing psychiatric treatment from the family physician.AimsOur objective is to understand the role of the family physician in help-seeking, recognition and ongoing management of first-episode psychosis.MethodWe will use a mixed-methods approach, incorporating health administrative data, electronic medical records (EMRs) and qualitative methodologies to study the role of the family physician at three points on the pathway to care. First, help-seeking: we will use health administrative data to examine access to a family physician and patterns of primary care use preceding the first diagnosis of psychosis; second, recognition: we will identify first-onset cases of psychosis in health administrative data, and look back at linked EMRs from primary care to define a risk profile for undetected cases; and third, management: we will examine service provision to identified patients through EMR data, including patterns of contacts, prescriptions and referrals to specialised care. We will then conduct qualitative interviews and focus groups with key stakeholders to better understand the trends observed in the quantitative data.DiscussionThese findings will provide an in-depth description of first-episode psychosis in primary care, informing strategies to build linkages between family physicians and psychiatric services to improve transitions of care during the crucial early stages of psychosis.Declaration of interestNone.

show abstract

“…However, in the practical scenarios, customers share similarly only on a small fraction of variables, such as knowledge, need, attitude, interest and loyalty status (Grover and Srinivasan, 1987;Yankelovich and Meer, 2006). In other words, these clustering methods obtain a global model rather than a local model, failing to discover subgroups of customers who have similar characteristics on partial variables, especially in high-dimensional data (Zhao et al, 2012).…”

Section: Introductionmentioning

confidence: 99%