OBJECTIVE: We aimed to develop a model for accurate prediction of general care inpatient deterioration. MATERIALS AND METHODS: Training and internal validation datasets were built using 2-year data from a quaternary hospital in the Midwest. Model training used gradient boosting and feature engineering (clinically relevant interactions, time-series information) to predict general care inpatient deterioration (resuscitation call, intensive care unit transfer, or rapid response team call) in 24 hours. Data from a tertiary care hospital in the Southwest were used for external validation. C-statistic, sensitivity, positive predictive value, and alert rate were calculated for different cutoffs and compared with the National Early Warning Score. Sensitivity analysis evaluated prediction of intensive care unit transfer or resuscitation call. RESULTS: Training, internal validation, and external validation datasets included 24 500, 25 784 and 53 956 hospitalizations, respectively. The Mayo Clinic Early Warning Score (MC-EWS) demonstrated excellent discrimination in both the internal and external validation datasets (C-statistic = 0.913, 0.937, respectively), and results were consistent in the sensitivity analysis (C-statistic = 0.932 in external validation). At a sensitivity of 73%, MC-EWS would generate 0.7 alerts per day per 10 patients, 45% less than the National Early Warning Score. DISCUSSION: Low alert rates are important for implementation of an alert system. Other early warning scores developed for the general care ward have achieved lower discrimination overall compared with MC-EWS, likely because MC-EWS includes both nursing assessments and extensive feature engineering. CONCLUSIONS: MC-EWS achieved superior prediction of general care inpatient deterioration using sophisticated feature engineering and a machine learning approach, reducing alert rate.

Early detection of patients with elevated risk of developing diabetes mellitus is critical to the improved prevention and overall clinical management of these patients. We aim to apply association rule mining to electronic medical records (EMR) to discover sets of risk factors and their corresponding subpopulations that represent patients at particularly high risk of developing diabetes. Given the high dimensionality of EMRs, association rule mining generates a very large set of rules which we need to summarize for easy clinical use. We reviewed four association rule set summarization techniques and conducted a comparative evaluation to provide guidance regarding their applicability, strengths and weaknesses. We proposed extensions to incorporate risk of diabetes into the process of finding an optimal summary. We evaluated these modified techniques on a real-world prediabetic patient cohort. We found that all four methods produced summaries that described subpopulations at high risk of diabetes with each method having its clear strength. For our purpose, our extension to the Buttom-Up Summarization (BUS) algorithm produced the most suitable summary. The subpopulations identified by this summary covered most high-risk patients, had low overlap and were at very high risk of diabetes.

Abstract Summary: Mate pair library sequencing is an effective and economical method for detecting genomic structural variants and chromosomal abnormalities. Unfortunately, the mapping and alignment of mate-pair read pairs to a reference genome is a challenging and time-consuming process for most next-generation sequencing alignment programs. Large insert sizes, introduction of library preparation protocol artifacts (biotin junction reads, paired-end read contamination, chimeras, etc.) and presence of structural variant breakpoints within reads increase mapping and alignment complexity. We describe an algorithm that is up to 20 times faster and 25% more accurate than popular next-generation sequencing alignment programs when processing mate pair sequencing. Availability: http://bioinformaticstools.mayo.edu/research/bima/ Contact: [email protected] or [email protected] Supplementary Information: Supplementary data are available at Bioinformatics online.

MOTIVATION: Batch effects are due to probe-specific systematic variation between groups of samples (batches) resulting from experimental features that are not of biological interest. Principal component analysis (PCA) is commonly used as a visual tool to determine whether batch effects exist after applying a global normalization method. However, PCA yields linear combinations of the variables that contribute maximum variance and thus will not necessarily detect batch effects if they are not the largest source of variability in the data. RESULTS: We present an extension of PCA to quantify the existence of batch effects, called guided PCA (gPCA). We describe a test statistic that uses gPCA to test whether a batch effect exists. We apply our proposed test statistic derived using gPCA to simulated data and to two copy number variation case studies: the first study consisted of 614 samples from a breast cancer family study using Illumina Human 660 bead-chip arrays, whereas the second case study consisted of 703 samples from a family blood pressure study that used Affymetrix SNP Array 6.0. We demonstrate that our statistic has good statistical properties and is able to identify significant batch effects in two copy number variation case studies. CONCLUSION: We developed a new statistic that uses gPCA to identify whether batch effects exist in high-throughput genomic data. Although our examples pertain to copy number data, gPCA is general and can be used on other data types as well. AVAILABILITY AND IMPLEMENTATION: The gPCA R package (Available via CRAN) provides functionality and data to perform the methods in this article. CONTACT: [email protected]

BACKGROUND: Microarray measurements are susceptible to a variety of experimental artifacts, some of which give rise to systematic biases that are spatially dependent in a unique way on each chip. It is likely that such artifacts affect many SNP arrays, but the normalization methods used in currently available genotyping algorithms make no attempt at spatial bias correction. Here, we propose an effective single-chip spatial bias removal procedure for Affymetrix 6.0 SNP arrays or platforms with similar design features. This procedure deals with both extreme and subtle biases and is intended to be applied before standard genotype calling algorithms. RESULTS: Application of the spatial bias adjustments on HapMap samples resulted in higher genotype call rates with equal or even better accuracy for thousands of SNPs. Consequently the normalization procedure is expected to lead to more meaningful biological inferences and could be valuable for genome-wide SNP analysis. CONCLUSIONS: Spatial normalization can potentially rescue thousands of SNPs in a genetic study at the small cost of computational time. The approach is implemented in R and available from the authors upon request.

A fuzzy c-means algorithm was adapted for analyzing microarray data. The adaptation consisted of initialization of fuzzy centroids using gene ontology information and the use of Pearson correlation distance in the objective function. To initialize fuzzy centroids, we classified genes based on gene ontology terms and used the classified genes as initial fuzzy clusters. Pearson correlation distance becomes 0 if two genes are either positively or negatively correlated. The algorithm was applied to Yeast and lung cancer microarray datasets. It outperformed the conventional fuzzy c-means algorithm by associating more genes to functional groups.

We describe a method to identify candidate cancer biomarkers by analyzing numeric approximations of tissue specificity of human genes. These approximations were calculated by analyzing predicted tissue expression distributions of genes derived from mapping expressed sequence tags (ESTs) to the human genome sequence using a binary indexing algorithm. Tissue-specificity values facilitated high-throughput analysis of the human genes and enabled the identification of genes highly specific to different tissues. Tissue expression distributions for several genes were compared to estimates obtained from other public gene expression datasets and experimentally validated using quantitative RT-PCR on RNA isolated from several human tissues. Our results demonstrate that most human genes ( approximately 98%) are expressed in many tissues (low specificity), and only a small number of genes possess very specific tissue expression profiles. These genes comprise a rich dataset from which novel therapeutic targets and novel diagnostic serum biomarkers may be selected.

MOTIVATION: Using stable isotopes in global proteome scans, labeled molecules from one sample are pooled with unlabeled molecules from another sample and subsequently subjected to mass-spectral analysis. Stable-isotope methodologies make use of the fact that identical molecules of different stable-isotope compositions are differentiated in a mass spectrometer and are represented in a mass spectrum as distinct isotopic clusters with a known mass shift. We describe two multivariable linear regression models for (16)O/(18)O stable-isotope labeled data that jointly model pairs of resolved isotopic clusters from the same peptide and quantify the abundance present in each of the two biological samples while concurrently accounting for peptide-specific incorporation rates of the heavy isotope. The abundance measure for each peptide from the two biological samples is then used in down-stream statistical analyses, e.g. differential expression analysis. Because the multivariable regression models are able to correct for the abundance of the labeled peptide that appear as an unlabeled peptide due to the inability to exchange the natural C-terminal oxygen for the heavy isotope, they are particularly advantageous for a two-step digestion/labeling procedure. We discuss how estimates from the regression model are used to quantify the variability of the estimated abundance measures for the paired samples. Although discussed in the context of (16)O/(18)O stable-isotope labeled data, the multivariable regression models are generalizable to other stable-isotope labeled technologies.

MOTIVATION: The need for normalization in microarray experiments has been well documented in the literature. Currently, many analysis methods treat normalization and analysis as a series of steps, with summarized data carried forward to the next step. RESULTS: We present a unified algorithm which incorporates normalization and class comparison in one analysis using probe level perfect match and mismatch data. The algorithm is based on calibration models common to most biological assays, and the resulting chip-specific parameters have a natural interpretation. We show that the algorithm fits into the statistical generalized linear models framework, describe a practical fitting strategy and present results of the algorithm applied to an example dataset as well as based on metrics used in affycomp. The algorithm ranks amongst the top third of the affycomp competitors, performing best in measures of bias. AVAILABILITY: R functions are available on request from the authors.

MOTIVATION: An important underlying assumption of any experiment is that the experimental subjects are similar across levels of the treatment variable, so that changes in the response variable can be attributed to exposure to the treatment under study. This assumption is often not valid in the analysis of a microarray experiment due to systematic biases in the measured expression levels related to experimental factors such as spot location (often referred to as a print-tip effect), arrays, dyes, and various interactions of these effects. Thus, normalization is a critical initial step in the analysis of a microarray experiment, where the objective is to balance the individual signal intensity levels across the experimental factors, while maintaining the effect due to the treatment under investigation. RESULTS: Various normalization strategies have been developed including log-median centering, analysis of variance modeling, and local regression smoothing methods for removing linear and/or intensity-dependent systematic effects in two-channel microarray experiments. We describe a method that incorporates many of these into a single strategy, referred to as two-channel fastlo, and is derived from a normalization procedure that was developed for single-channel arrays. The proposed normalization procedure is applied to a two-channel dose-response experiment.

MOTIVATION: Our goal was to develop a normalization technique that yields results similar to cyclic loess normalization and with speed comparable to quantile normalization. RESULTS: Fastlo yields normalized values similar to cyclic loess and quantile normalization and is fast; it is at least an order of magnitude faster than cyclic loess and approaches the speed of quantile normalization. Furthermore, fastlo is more versatile than both cyclic loess and quantile normalization because it is model-based. AVAILABILITY: The Splus/R function for fastlo normalization is available from the authors.