Unmixing biological fluorescence image data with sparse and low-rank Poisson regression.

MOTIVATION: Multispectral biological fluorescence microscopy has enabled the identification of multiple targets in complex samples. The accuracy in the unmixing result degrades (i) as the number of fluorophores used in any experiment increases and (ii) as the signal-to-noise ratio in the recorded images decreases. Further, the availability of prior knowledge regarding the expected spatial distributions of fluorophores in images of labeled cells provides an opportunity to improve the accuracy of fluorophore identification and abundance. RESULTS: We propose a regularized sparse and low-rank Poisson regression unmixing approach (SL-PRU) to deconvolve spectral images labeled with highly overlapping fluorophores which are recorded in low signal-to-noise regimes. First, SL-PRU implements multipenalty terms when pursuing sparseness and spatial correlation of the resulting abundances in small neighborhoods simultaneously. Second, SL-PRU makes use of Poisson regression for unmixing instead of least squares regression to better estimate photon abundance. Third, we propose a method to tune the SL-PRU parameters involved in the unmixing procedure in the absence of knowledge of the ground truth abundance information in a recorded image. By validating on simulated and real-world images, we show that our proposed method leads to improved accuracy in unmixing fluorophores with highly overlapping spectra. AVAILABILITY AND IMPLEMENTATION: The source code used for this article was written in MATLAB and is available with the test data at https://github.com/WANGRUOGU/SL-PRU.

Unmixing Biological Fluorescence Image Data with Sparse and Low-Rank Poisson Regression.

Multispectral biological fluorescence microscopy has enabled the identification of multiple targets in complex samples. The accuracy in the unmixing result degrades (1) as the number of fluorophores used in any experiment increases and (2) as the signal-to-noise ratio in the recorded images decreases. Further, the availability of prior knowledge regarding the expected spatial distributions of fluorophores in images of labeled cells provides an opportunity to improve the accuracy of fluorophore identification and abundance. We propose a regularized sparse and low-rank Poisson unmixing approach (SL-PRU) to deconvolve spectral images labeled with highly overlapping fluorophores which are recorded in low signal-to-noise regimes. Firstly, SL-PRU implements multi-penalty terms when pursuing sparseness and spatial correlation of the resulting abundances in small neighborhoods simultaneously. Secondly, SL-PRU makes use of Poisson regression for unmixing instead of least squares regression to better estimate photon abundance. Thirdly, we propose a method to tune the SL-PRU parameters involved in the unmixing procedure in the absence of knowledge of the ground truth abundance information in a recorded image. By validating on simulated and real-world images, we show that our proposed method leads to improved accuracy in unmixing fluorophores with highly overlapping spectra.

High Proton-Conductive and Temperature-Tolerant PVC-P4VP Membranes towards Medium-Temperature Water Electrolysis.

Water electrolysis (WE) is a highly promising approach to producing clean hydrogen. Medium-temperature WE (100-350 °C) can improve the energy efficiency and utilize the low-grade water vapor. Therefore, a high-temperature proton-conductive membrane is desirable to realize the medium-temperature WE. Here, we present a polyvinyl chloride (PVC)-poly(4vinylpyridine) (P4VP) hybrid membrane by a simple cross-linking of PVC and P4VP. The pyridine groups of P4VP promote the loading rate of phosphoric acid, which delivers the proton conductivity of the PVC-P4VP membrane. The optimized PVC-P4VP membrane with a 1:2 content ratio offers the maximum proton conductivity of 4.3 × 10-2 S cm-1 at 180 °C and a reliable conductivity stability in 200 h at 160 °C. The PVC-P4VP membrane electrode is covered by an IrO2 anode, and a Pt/C cathode delivers not only the high water electrolytic reactivity at 100-180 °C but also the stable WE stability at 180 °C.

Average Top-k Aggregate Loss for Supervised Learning.

In this work, we introduce the average top- k ( ATk) loss, which is the average over the k largest individual losses over a training data, as a new aggregate loss for supervised learning. We show that the ATk loss is a natural generalization of the two widely used aggregate losses, namely the average loss and the maximum loss. Yet, the ATk loss can better adapt to different data distributions because of the extra flexibility provided by the different choices of k. Furthermore, it remains a convex function over all individual losses and can be combined with different types of individual loss without significant increase in computation. We then provide interpretations of the ATk loss from the perspective of the modification of individual loss and robustness to training data distributions. We further study the classification calibration of the ATk loss and the error bounds of ATk-SVM model. We demonstrate the applicability of minimum average top- k learning for supervised learning problems including binary/multi-class classification and regression, using experiments on both synthetic and real datasets.

Local Infiltration Analgesia Versus Regional Blockade for Postoperative Analgesia in Total Knee Arthroplasty: A Meta-analysis of Randomized Controlled Trials.

BACKGROUND: Total knee arthroplasty (TKA) is one of the most commonly performed procedures while postoperative analgesia still remains challenging. The efficacy and safety of local infiltration analgesia (LIA) versus regional blockade (RB; epidural analgesia and/or peripheral nerve block) for pain management after TKA are controversial. OBJECTIVES: The purpose of this meta-analysis was to determine whether LIA compared with RB would provide better postoperative pain control, consume less morphine, facilitate early functional recovery, entail a differential risk of side effects and complications, and allow a shorter length of stay. STUDY DESIGN: This meta-analysis pooled all data published in randomized controlled trials (RCTs) examining the efficacy and safety of LIA versus RB following TKA. SETTING: The work was performed at Affiliated Cixi Hospital, Wenzhou Medical University. METHODS: Literature in English was searched using EMBASE, Medline, Cochrane Library, CINAHL, Web of Science, and Scopus from inception to April 2015. RCTs that compared LIA and RB for postoperative analgesia following TKA were included. Methodological quality was assessed using the Cochrane Back Review Group checklist, and a sensitivity analysis was performed. Sixteen RCTs with a total of 1,206 patients were finally included in our study. RESULTS: The results of our meta-analysis indicate that patients managed by LIA showed significantly lower numeric rating scale (NRS) score at rest (WMD: -0.40 [-0.72, -0.07]; P = 0.02) when compared with those managed by RB. Difference of morphine consumption was not significant (WMD: -1.39 [-7.21, 4.44]; P = 0.64) between the 2 groups. In terms of early functional recovery, the LIA group showed more straight leg raise (RR: 2.90 [2.15, 3.93]; P < 0.00001) on the first postoperative day; better range of motion within one week (WMD: 4.33 [2.61, 6.05]; P < 0.00001), but not at 3 months (WMD: 1.98 [-0.02, 3.98]; P = 0.05); and comparable knee society score (WMD: -8.79 [-27.05, 9.48]; P = 0.35). Length of hospital stay of the LIA group was marginally shorter (WMD: -0.25 [-0.49, -0.01]; P = 0.05) than that of the RB group. Risk of side effects and complications were comparable between groups. LIMITATIONS: The lack of a standard criterion regarding the technique details of LIA and heterogeneity resulting from the various analgesic components, dosages, and different administration methods might have posed a bias on the results. CONCLUSION: Our results have indicated that LIA provided better analgesia than RB at rest and preserved quadriceps function in the immediate postoperative period, which may be beneficial to early functional recovery. And its safety profile is reliable. With the biases in our meta-analysis, a rigorous and adequately powered RCT is needed to validate our results. KEY WORDS: Local infiltration analgesia, regional block, peripheral nerve block, epidural analgesia, postoperative analgesia, total knee arthroplasty, meta-analysis, randomized controlled trial.

Online Pairwise Learning Algorithms.

Pairwise learning usually refers to a learning task that involves a loss function depending on pairs of examples, among which the most notable ones are bipartite ranking, metric learning, and AUC maximization. In this letter we study an online algorithm for pairwise learning with a least-square loss function in an unconstrained setting of a reproducing kernel Hilbert space (RKHS) that we refer to as the Online Pairwise lEaRning Algorithm (OPERA). In contrast to existing works (Kar, Sriperumbudur, Jain, & Karnick, 2013 ; Wang, Khardon, Pechyony, & Jones, 2012 ), which require that the iterates are restricted to a bounded domain or the loss function is strongly convex, OPERA is associated with a non-strongly convex objective function and learns the target function in an unconstrained RKHS. Specifically, we establish a general theorem that guarantees the almost sure convergence for the last iterate of OPERA without any assumptions on the underlying distribution. Explicit convergence rates are derived under the condition of polynomially decaying step sizes. We also establish an interesting property for a family of widely used kernels in the setting of pairwise learning and illustrate the convergence results using such kernels. Our methodology mainly depends on the characterization of RKHSs using its associated integral operators and probability inequalities for random variables with values in a Hilbert space.

Probabilistic inference of biological networks via data integration.

There is significant interest in inferring the structure of subcellular networks of interaction. Here we consider supervised interactive network inference in which a reference set of known network links and nonlinks is used to train a classifier for predicting new links. Many types of data are relevant to inferring functional links between genes, motivating the use of data integration. We use pairwise kernels to predict novel links, along with multiple kernel learning to integrate distinct sources of data into a decision function. We evaluate various pairwise kernels to establish which are most informative and compare individual kernel accuracies with accuracies for weighted combinations. By associating a probability measure with classifier predictions, we enable cautious classification, which can increase accuracy by restricting predictions to high-confidence instances, and data cleaning that can mitigate the influence of mislabeled training instances. Although one pairwise kernel (the tensor product pairwise kernel) appears to work best, different kernels may contribute complimentary information about interactions: experiments in S. cerevisiae (yeast) reveal that a weighted combination of pairwise kernels applied to different types of data yields the highest predictive accuracy. Combined with cautious classification and data cleaning, we can achieve predictive accuracies of up to 99.6%.

Guaranteed classification via regularized similarity learning.

Learning an appropriate (dis)similarity function from the available data is a central problem in machine learning, since the success of many machine learning algorithms critically depends on the choice of a similarity function to compare examples. Despite many approaches to similarity metric learning that have been proposed, there has been little theoretical study on the links between similarity metric learning and the classification performance of the resulting classifier. In this letter, we propose a regularized similarity learning formulation associated with general matrix norms and establish their generalization bounds. We show that the generalization error of the resulting linear classifier can be bounded by the derived generalization bound of similarity learning. This shows that a good generalization of the learned similarity function guarantees a good classification of the resulting linear classifier. Our results extend and improve those obtained by Bellet, Habrard, and Sebban (2012). Due to the techniques dependent on the notion of uniform stability (Bousquet & Elisseeff, 2002), the bound obtained there holds true only for the Frobenius matrix-norm regularization. Our techniques using the Rademacher complexity (Bartlett & Mendelson, 2002) and its related Khinchin-type inequality enable us to establish bounds for regularized similarity learning formulations associated with general matrix norms, including sparse L1-norm and mixed (2,1)-norm.

Rademacher chaos complexities for learning the kernel problem.

We develop a novel generalization bound for learning the kernel problem. First, we show that the generalization analysis of the kernel learning problem reduces to investigation of the suprema of the Rademacher chaos process of order 2 over candidate kernels, which we refer to as Rademacher chaos complexity. Next, we show how to estimate the empirical Rademacher chaos complexity by well-established metric entropy integrals and pseudo-dimension of the set of candidate kernels. Our new methodology mainly depends on the principal theory of U-processes and entropy integrals. Finally, we establish satisfactory excess generalization bounds and misclassification error rates for learning gaussian kernels and general radial basis kernels.

Enhanced protein fold recognition through a novel data integration approach.

BACKGROUND: Protein fold recognition is a key step in protein three-dimensional (3D) structure discovery. There are multiple fold discriminatory data sources which use physicochemical and structural properties as well as further data sources derived from local sequence alignments. This raises the issue of finding the most efficient method for combining these different informative data sources and exploring their relative significance for protein fold classification. Kernel methods have been extensively used for biological data analysis. They can incorporate separate fold discriminatory features into kernel matrices which encode the similarity between samples in their respective data sources. RESULTS: In this paper we consider the problem of integrating multiple data sources using a kernel-based approach. We propose a novel information-theoretic approach based on a Kullback-Leibler (KL) divergence between the output kernel matrix and the input kernel matrix so as to integrate heterogeneous data sources. One of the most appealing properties of this approach is that it can easily cope with multi-class classification and multi-task learning by an appropriate choice of the output kernel matrix. Based on the position of the output and input kernel matrices in the KL-divergence objective, there are two formulations which we respectively refer to as MKLdiv-dc and MKLdiv-conv. We propose to efficiently solve MKLdiv-dc by a difference of convex (DC) programming method and MKLdiv-conv by a projected gradient descent algorithm. The effectiveness of the proposed approaches is evaluated on a benchmark dataset for protein fold recognition and a yeast protein function prediction problem. CONCLUSION: Our proposed methods MKLdiv-dc and MKLdiv-conv are able to achieve state-of-the-art performance on the SCOP PDB-40D benchmark dataset for protein fold prediction and provide useful insights into the relative significance of informative data sources. In particular, MKLdiv-dc further improves the fold discrimination accuracy to 75.19% which is a more than 5% improvement over competitive Bayesian probabilistic and SVM margin-based kernel learning methods. Furthermore, we report a competitive performance on the yeast protein function prediction problem.

Bayesian unsupervised learning with multiple data types.

We propose Bayesian generative models for unsupervised learning with two types of data and an assumed dependency of one type of data on the other. We consider two algorithmic approaches, based on a correspondence model, where latent variables are shared across datasets. These models indicate the appropriate number of clusters in addition to indicating relevant features in both types of data. We evaluate the model on artificially created data. We then apply the method to a breast cancer dataset consisting of gene expression and microRNA array data derived from the same patients. We assume partial dependence of gene expression on microRNA expression in this study. The method ranks genes within subtypes which have statistically significant abnormal expression and ranks associated abnormally expressing microRNA. We report a genetic signature for the basal-like subtype of breast cancer found across a number of previous gene expression array studies. Using the two algorithmic approaches we find that this signature also arises from clustering on the microRNA expression data and appears derivative from this data.

A marginalized variational bayesian approach to the analysis of array data.

BACKGROUND: Bayesian unsupervised learning methods have many applications in the analysis of biological data. For example, for the cancer expression array datasets presented in this study, they can be used to resolve possible disease subtypes and to indicate statistically significant dysregulated genes within these subtypes. RESULTS: In this paper we outline a marginalized variational Bayesian inference method for unsupervised clustering. In this approach latent process variables and model parameters are allowed to be dependent. This is achieved by marginalizing the mixing Dirichlet variables and then performing inference in the reduced variable space. An iterative update procedure is proposed. CONCLUSION: Theoretically and experimentally we show that the proposed algorithm gives a much better free-energy lower bound than a standard variational Bayesian approach. The algorithm is computationally efficient and its performance is demonstrated on two expression array data sets.