Pitfalls in the use of DNA microarray data for diagnostic and prognostic classification.

Journal of the National Cancer Institute, 2003

737 citations recorded by the source corpus.

Subfields: Molecular Biology

View Paper: https://www.semanticscholar.org/paper/bdd7686f978c075e0229bf5f5ade976636cd06ed

Citations in the corpus

• Prognostic gene signatures for non-small-cell lung cancer (2009, neutral): “We also performed a preliminary validation by using leave-one-out cross-validation (24).”
• Sparse evolutionary deep learning with over one million artificial neurons on commodity hardware (2019, positive): “Especially, gene expression obtained from DNA microarray has emerged as a powerful solution to cure cancers (Simon et al., 2003).”
• Clinical biomarkers in drug discovery and development (2003, positive): “However, attention must be paid to protocol design and analysis if confounders are to be managed and these huge volumes of data are to be distilled into useful bioinformatic”
• A comprehensive evaluation of multicategory classification methods for microbiomic data (2013, neutral): “We focused here on supervised classification methods because unsupervised methods (such as clustering and principal component analysis), which are designed to reveal structure of the data, provide visual summaries, and help quality control, are not optimal (and depending on application may also be c”
• Toward more realistic drug–target interaction predictions (2014, positive): “Another source of potential bias is that simple cross-validation (CV) cannot evaluate the effect of adjusting the model parameters, and may therefore easily lead to selection bias and overoptimistic prediction results [20–22].”
• GEMS: A system for automated cancer diagnosis and biomarker discovery from microarray gene expression data (2005, neutral): “Since prior research has demonstrated superiority of supervised classification methods for cancer diagnosis over unsupervised techniques [7], we focused only on systems implementing supervised classification algorithms.”
• Bias in error estimation when using cross-validation for model selection (2006, neutral): “[2], Ambroise and McLachlan [3] and Reunanen [4], this gives a very biased estimate of the true error; not much better than the resubstitution estimate.”
• Prediction of survival in diffuse large-B-cell lymphoma based on the expression of six genes. (2004, neutral): “TTG-2, a new gene encoding a cysteine-rich protein with the LIM motif, is overexpressed in acute T-cell leukaemia with the t(11;14)(p13;q11).”
• Predictive data mining in clinical medicine: Current issues and guidelines (2008, neutral): “[57].”
• Classifying Gene Expression Profiles from Pairwise mRNA Comparisons (2004, positive): “The small-sample dilemma in the statistical analysis of microarray data is well-documented in the literature (Dudoit and Fridlyand (2003); Sebastiani et al. (2003); Simon et al. (2003); West et al. (2001)).”
• Molecular classification and molecular forecasting of breast cancer: ready for clinical application? (2005, positive): “Because these methods do not derive statistically valid information regarding which genes are differentially expressed between classes, they are poorly suited for identifying prognostic variables.(37) The aim of supervised classification of data is to detect genes that may predict defined outcomes.”
• Applied Predictive Modeling (2013, neutral): “Over-fitting has been discussed in the fields of forecasting (Clark 2004), medical research (Simon et al. 2003; Steyerberg 2010), chemometrics (Gowen et al.”
• Prediction error estimation: a comparison of resampling methods (2005, positive): “Feature selection must be based on the learning set within each resampling, otherwise additional bias is introduced (Simon et al., 2003).”
• Gene Expression-Based Classification of Non-Small Cell Lung Carcinomas and Survival Prediction (2010, positive): “The predictive value of the prognosis model was evaluated by performing ‘‘leave-one-out’’ cross-validation [16,17].”