Showing papers by "Goutam Saha published in 2016"

Construction of Gene Regulatory Networks Using Recurrent Neural Networks and Swarm Intelligence.

Abstract: We have proposed a methodology for the reverse engineering of biologically plausible gene regulatory networks from temporal genetic expression data. We have used established information and the fundamental mathematical theory for this purpose. We have employed the Recurrent Neural Network formalism to extract the underlying dynamics present in the time series expression data accurately. We have introduced a new hybrid swarm intelligence framework for the accurate training of the model parameters. The proposed methodology has been first applied to a small artificial network, and the results obtained suggest that it can produce the best results available in the contemporary literature, to the best of our knowledge. Subsequently, we have implemented our proposed framework on experimental (in vivo) datasets. Finally, we have investigated two medium sized genetic networks (in silico) extracted from GeneNetWeaver, to understand how the proposed algorithm scales up with network size. Additionally, we have implemented our proposed algorithm with half the number of time points. The results indicate that a reduction of 50% in the number of time points does not have an effect on the accuracy of the proposed methodology significantly, with a maximum of just over 15% deterioration in the worst case.

Reverse engineering of gene regulatory networks based on S-systems and Bat algorithm.

Abstract: The correct inference of gene regulatory networks for the understanding of the intricacies of the complex biological regulations remains an intriguing task for researchers. With the availability of large dimensional microarray data, relationships among thousands of genes can be simultaneously extracted. Among the prevalent models of reverse engineering genetic networks, S-system is considered to be an efficient mathematical tool. In this paper, Bat algorithm, based on the echolocation of bats, has been used to optimize the S-system model parameters. A decoupled S-system has been implemented to reduce the complexity of the algorithm. Initially, the proposed method has been successfully tested on an artificial network with and without the presence of noise. Based on the fact that a real-life genetic network is sparsely connected, a novel Accumulative Cardinality based decoupled S-system has been proposed. The cardinality has been varied from zero up to a maximum value, and this model has been implemented for the reconstruction of the DNA SOS repair network of Escherichia coli. The obtained results have shown significant improvements in the detection of a greater number of true regulations, and in the minimization of false detections compared to other existing methods.

Large-Scale Recurrent Neural Network Based Modelling of Gene Regulatory Network Using Cuckoo Search-Flower Pollination Algorithm

Abstract: The accurate prediction of genetic networks using computational tools is one of the greatest challenges in the postgenomic era. Recurrent Neural Network is one of the most popular but simple approaches to model the network dynamics from time-series microarray data. To date, it has been successfully applied to computationally derive small-scale artificial and real-world genetic networks with high accuracy. However, they underperformed for large-scale genetic networks. Here, a new methodology has been proposed where a hybrid Cuckoo Search-Flower Pollination Algorithm has been implemented with Recurrent Neural Network. Cuckoo Search is used to search the best combination of regulators. Moreover, Flower Pollination Algorithm is applied to optimize the model parameters of the Recurrent Neural Network formalism. Initially, the proposed method is tested on a benchmark large-scale artificial network for both noiseless and noisy data. The results obtained show that the proposed methodology is capable of increasing the inference of correct regulations and decreasing false regulations to a high degree. Secondly, the proposed methodology has been validated against the real-world dataset of the DNA SOS repair network of Escherichia coli. However, the proposed method sacrifices computational time complexity in both cases due to the hybrid optimization process.

Multiple parallel assay operations with cross contamination avoidance in a given biochip

Abstract: Digital microfluidic biochips are reforming many areas of biochemistry, biomedical sciences, as well as microelectronics. It is renowned as lab-on-a-chip for its appreciation as a substitute for laboratory experiments. Nowadays, for emergency purposes and to ensure cost efficacy, multiple assay operations are essential to be carried out simultaneously. In this context, parallelism is of utmost importance in designing biochip while the size of a chip is a constraint. Hence, the objective of this study is to enhance the performance of a chip in terms of its throughput, electrode utilisation, and pin count as well. Here, the authors have considered some of the most familiar assay requirements where a sample is to be analysed using different reagents, and identify some parameter(s) of the sample(s) under consideration. Moreover, sample preparation is a vital task in digital microfluidic biochip; thus, dilution of different samples up to different concentrations using buffer (neutral) fluid is a crucial issue. In this design, the authors effectively perform this task in parallel in a number of sub-regions of a given restricted sized chip using an array based partitioning pin-assignment technique while taking care of the cross contamination problem. The design has been verified for some significant real life assay examples.

A Survey on Recurrent Neural Network Based Modelling of Gene Regulatory Network

Abstract: The correct inference of gene regulatory networks (GRN) remains as a fascinating task for researchers to understand the detailed process of complex biological regulations and functions. With availability of large dimensional microarray data, relationships among thousands of genes can be extracted simultaneously that is a reverse engineering problem. Among the different popular models to infer GRN, Recurrent Neural Networks (RNN) are considered as most popular and promising mathematical tool to model the dynamics of, as well as to infer the correct dependencies among genes from, biological data like time series microarray. RNN is closed loop Neural Network with a delay feedback. By observing the weights of RNN model, it is possible to extract the regulations among genes. Several metaheuristics or optimization techniques were already proposed to search the optimal value of RNN model parameters. In this review, we illustrate different problems regarding reverse engineering of GRN and how different proposed models can overcome these problems. It is observed that finding out the most suitable and efficient optimization techniques for the accurate inference of small artificial, large artificial, Dream4 Network,and real world GRNs with less computational complexity are still an open research problem to all.

Inference of Gene Regulatory Networks with Neural-Cuckoo Hybrid

Abstract: Current progress in cellular biology and bioinformatics allow researchers to get a distinct picture of the complex biochemical processes those occur within a cell of the human body and remain as the cause for many diseases. Therefore, this technology opened up a new door to the researchers of computer science as well as to biologists to work together to investigate the causes of a disease. One of the greatest challenges of the post-genomic era is the investigation and inference of the regulatory interactions or dependencies between genes from the microarray data. Here, a new methodology has been devised for investigating the genetic interactions among genes from temporal gene expression data by combining the features of Neural Network and Cuckoo Search optimization. The developed technique has been applied on the real-world microarray dataset of Lung Adenocarcinoma for detection of genes which may be directly responsible for the cause of Lung Adenocarcinoma.

Robustness of Voice Conversion Techniques Under Mismatched Conditions

Abstract: Most of the existing studies on voice conversion (VC) are conducted in acoustically matched conditions between source and target signal. However, the robustness of VC methods in presence of mismatch remains unknown. In this paper, we report a comparative analysis of different VC techniques under mismatched conditions. The extensive experiments with five different VC techniques on CMU ARCTIC corpus suggest that performance of VC methods substantially degrades in noisy conditions. We have found that bilinear frequency warping with amplitude scaling (BLFWAS) outperforms other methods in most of the noisy conditions. We further explore the suitability of different speech enhancement techniques for robust conversion. The objective evaluation results indicate that spectral subtraction and log minimum mean square error (logMMSE) based speech enhancement techniques can be used to improve the performance in specific noisy conditions.

A swarm intelligence based scheme for reduction of false positives in inferred gene regulatory networks

Abstract: A gene regulatory network reveals the regulatory relationships among genes at a cellular level. The accurate reconstruction of such networks using computational tools, from time series genetic expression data, is crucial to the understanding of the proper functioning of a living organism. Investigations in this domain focused mainly on the identification of as many true regulations as possible. This has somewhat overshadowed the reduction of false predictions in inferred networks. In the present investigation, we have proposed a novel scheme, based on different swarm intelligence algorithms, to reduce the number of inferred false regulations. We have first applied our proposed methodology on the much studied, benchmark experimental datasets of the DNA SOS repair network of Escherichia Coli. Subsequently, we have experimented upon a larger, in silico network extracted from the GeneNetWeaver database. The obtained results suggest that the proposed methodology can reduce the number of false predictions, significantly, without using any supplementary biological information for larger gene regulatory networks.

Smartphone based blood flow feature extraction and classification from Doppler spectrum images

Abstract: Ultrasound (US) Doppler spectrograms have been widely used for diagnosing vascular obstructions. This paper presents an Android smartphone based new approach for detecting the blood flow condition based on the US Doppler spectrogram images. A set of 59 spectrograms acquired from a US Doppler system is processed to extract features, and these non-redundant features are fed into a supervised classifier to determine the normal and abnormal blood flow. The classification is performed using the k-nearest neighbors (k-NN), Support vector machine (SVM), Naive Bayes (NB) and Multilayer perception (MLP) based classifiers. The SVM based classifier has shown superior performance, having an accuracy of 86.4 %, with a sensitivity and specificity of 96.4 % and 77.4 % respectively. The complete technique is implemented as an Android application and the results show the efficacy of the presented approach for the automated diagnosis of arterial diseases.