Our Publications

Panels of mRNAs and miRNAs for decoding molecular mechanisms of Renal Cell Carcinoma (RCC) subtypes utilizing Artificial Intelligence approaches

Renal Cell Carcinoma (RCC) encompasses three histological subtypes, including clear cell RCC (KIRC), papillary RCC (KIRP), and chromophobe RCC (KICH) each of which has different clinical courses, genetic/epigenetic drivers, and therapeutic responses. This study aimed to identify the significant mRNAs and microRNA panels involved in the pathogenesis of RCC subtypes. The mRNA and microRNA transcripts profile were obtained from The Cancer Genome Atlas (TCGA), which were included 611 ccRCC patients, 321 pRCC patients, and 89 chRCC patients for mRNA data and 616 patients in the ccRCC subtype, 326 patients in the pRCC subtype, and 91 patients in the chRCC for miRNA data, respectively. To identify mRNAs and miRNAs, feature selection based on filter and graph algorithms was applied. Then, a deep model was used to classify the subtypes of the RCC. Finally, an association rule mining algorithm was used to disclose features with significant roles to trigger molecular mechanisms to cause RCC subtypes. Panels of 77 mRNAs and 73 miRNAs could discriminate the KIRC, KIRP, and KICH subtypes from each other with 92% (F1-score ≥ 0.9, AUC ≥ 0.89) and 95% accuracy (F1-score ≥ 0.93, AUC ≥ 0.95), respectively. The Association Rule Mining analysis could identify miR-28 (repeat count = 2642) and CSN7A (repeat count = 5794) along with the miR-125a (repeat count = 2591) and NMD3 (repeat count = 2306) with the highest repeat counts, in the KIRC and KIRP rules, respectiv…

A self-organizing deep neuro-fuzzy system approach for classification of kidney cancer subtypes using miRNA genomics data

Kidney cancer is a dangerous disease affecting many patients all over the world. Early-stage diagnosis and correct identification of kidney cancer subtypes play an essential role in the patient's survival; therefore, its subtypes diagnosis and classification are the main challenges in kidney cancer treatment. Medical studies have proved that miRNA dysregulation can increase the risk of cancer. Thus, in this paper, we propose a new machine learning approach for significant miRNAs identification and kidney cancer subtype classification to design an automatic diagnostic tool. The proposed method contains two main steps: feature selection and classification. First, we apply the feature selection algorithm to choose the candidate miRNAs for each subtype. The feature selection algorithm utilizes the AMGM measure to select significant miRNAs with high discriminant power. Next, the candidate miRNAs are fed to a classifier to evaluate the candidate features. In the classification step, the proposed self-organizing deep neuro-fuzzy system is employed to classify kidney cancer subgroups. The new deep neuro-fuzzy system consists of a deep structure in the rule layer and novel architecture in the fuzzifier layer. The proposed self-organizing deep neuro-fuzzy system can help us to overcome the main obstacles in the field of neuro-fuzzy system applications, such as the curse of dimensionality. The goal of this paper is to illustrate that the neuro-fuzzy sys…

A Self-organizing Deep Auto-Encoder approach for Classification of Complex Diseases using SNP Genomics Data

Recently, many Machine Learning algorithms have been utilized to identify significant Single Nucleotide Polymorphisms (SNPs) in various human diseases. However, some principal obstacles are challenging in the field of SNP detection and healthy-patient classification. The curse of dimensionality is the main challenge. On the other hand, the number of samples is decidedly smaller than the number of SNPs. In addition, the number of healthy and patient samples can be unequal. These challenges make the feature selection and classification very difficult. The main goal of the current study is the combination of the various algorithms to find out the most effective way of SNP data analysis. Therefore, an efficient method is proposed to identify significant SNPs and classify healthy and patient samples. In this regard, firstly, the Mean Encoding, as an intelligent method, is utilized to convert the nominal SNP data to numeric. Then a two-step filter method is used for feature selection, which removes the irrelevant and redundant features. Finally, the proposed deep auto-encoder is employed to classify so that it can construct its structure based on input data, automatically. To evaluate, we apply the proposed approach to five different SNP datasets, including thyroid cancer, mental retardation, breast cancer, colorectal cancer, and autism, which obtained from the Gene Expression Omnibus (GEO) dataset. The proposed method has succeeded in feature sele…

Machine learning as new promising technique for selection of significant features in obese women with type 2 diabetes

Background

The global trend of obesity and diabetes is considerable. Recently, the early diagnosis and accurate prediction of type 2 diabetes mellitus (T2DM) patients have been planned to be estimated according to precise and reliable methods, artificial networks and machine learning (ML).

Materials and methods

In this study, an experimental data set of relevant features (adipocytokines and anthropometric levels) obtained from obese women (diabetic and non-diabetic) was analyzed. Machine learning was used to select significant features [by the separability-correlation measure (SCM) algorithm] for classification of women with the best accuracy and the results were evaluated using an artificial neural network (ANN).

Results

According to the experimental data analysis, a significant difference (p < 0.05) was found between fasting blood sugar (FBS), hemoglobin A1c (HbA1c) and visfatin level in two groups. Moreover, significant correlations were determined between HbA1c and FBS, homeostatic model assessment (HOMA) and insulin, total cholesterol (TC) level and body mass index (BMI) in non-diabetic women and insulin and HOMA, FBS and HbA1c, insulin and HOMA, systolic blood pressure (SBP) and diastolic blood pressure (DBP), BMI and TC and HbA1c and TC in the diabetic gro…

The Self-Organizing Restricted Boltzmann Machine for Deep Representation with the Application on Classification Problems

Recently, deep learning is proliferating in the field of representation learning. A deep belief network (DBN) consists of a deep network architecture that can generate multiple features of input patterns, using restricted Boltzmann machines (RBMs) as a building block of DBN. A deep learning model can achieve extremely high accuracy in many applications that depend on the model structure. However, specifying various parameters of deep network architecture like the number of hidden layers and neurons is a difficult task even for expert designers. Besides, the number of hidden layers and neurons is typically set manually, while this method is costly in terms of time and computational cost, especially in big data. In this paper, we introduce an approach to determine the number of hidden layers and neurons of the deep network automatically during the learning process. To this end, the input vector is transformed from the feature space with a low dimension into the new feature space with a high dimension in a hidden layer of RBM. In the following, new features are ranked according to their discrimination power between classes in the new space, using the Separability-correlation measure for feature importance ranking algorithm. The algorithm uses the mean of weights as a threshold, so the neurons whose weights exceed the threshold are retained, and the others are removed in the hidden layer. The number of retained neurons is presented as a reasonabl…