A random forest model for predicting exosomal proteins using evolutionary information and motifs

Arora, Akanksha; Patiyal, Sumeet; Sharma, Neelam; Naorem, Leimarembi Devi; Kaur, Dashleen; Raghava, Gajendra P. S.

doi:10.1101/2023.01.30.526378

Cited by 5 publications

(10 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The protein’s evolutionary features are known to provide additional crucial information about proteins [59,60]. The evolutionary information of proteins was retrieved from a position-specific scoring matrix (PSSM) profile generated using Position-Specific Iterated BLAST (PSI-BLAST) [61].…”

Section: Methodsmentioning

confidence: 99%

“…It has been shown in number of studies in the past that evolutionary information based models perform better than single sequence models [72][73][74][75]. Thus, this study also explores the potential of evolutionary information in predicting AFPs.…”

Section: Pssm Profilesmentioning

confidence: 97%

“…We devised motifs from exclusive and inclusive motifs by determining values such as a) No gap, b) Gap = 1, c) Gap = 2, and d) Class = Koolman-Rohm. Following that, the unique proteins possessing the motifs were identified to estimate the total coverage of motifs in the protein sequence [60].…”

Section: Motif Based Featuresmentioning

confidence: 99%

“…The model was trained and evaluated based on five-fold cross-validation and the independent validation dataset [51]. Finally, the mean of the performances of five iterations is used as the overall performance [60,67].…”

Section: Cross-validation Techniquementioning

confidence: 99%

See 3 more Smart Citations

Prediction of anti-freezing proteins from their evolutionary profile

Kumar,

Choudhury,

Bajiya

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

Prediction of antifreeze proteins (AFPs) holds significant importance due to their diverse applications in healthcare. An inherent limitation of current AFP prediction methods is their reliance on unreviewed proteins for evaluation. This study evaluates proposed and existing methods on an independent dataset containing 81 AFPs and 73 non-AFPs obtained from Uniport, which have been already reviewed by experts. Initially, we constructed machine learning models for AFP prediction using selected composition-based protein features and achieved a peak AUC of 0.90 with an MCC of 0.69 on the independent dataset. Subsequently, we observed a notable enhancement in model performance, with the AUC increasing from 0.90 to 0.93 upon incorporating evolutionary information instead of relying solely on the primary sequence of proteins. Furthermore, we explored hybrid models integrating our machine learning approaches with BLAST-based similarity and motif-based methods. However, the performance of these hybrid models either matched or was inferior to that of our best machine-learning model. Our best model based on evolutionary information outperforms all existing methods on independent/validation dataset. To facilitate users, a user-friendly web server with a standalone package named AFPropred was developed (https://webs.iiitd.edu.in/raghava/afpropred).

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Pssm Profilesmentioning

confidence: 97%

Section: Motif Based Featuresmentioning

confidence: 99%

Section: Cross-validation Techniquementioning

confidence: 99%

See 2 more Smart Citations

Prediction of anti-freezing proteins from their evolutionary profile

Kumar,

Choudhury,

Bajiya

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…For eg., if a query sequence was hit against the hormonal peptide sequence, it was classified as hormone otherwise it was classified as non-hormone. Several researchers have used and thoroughly annotated this methodology [24,26]. The BLAST database was created using all sequences present in the training dataset.…”

Section: Blast For Similarity Searchmentioning

confidence: 99%

Prediction of peptide hormones using an ensemble of machine learning and similarity-based methods

Kaur

Arora

Vigneshwar

et al. 2023

Preprint

View full text Add to dashboard Cite

Peptide hormones are genome-encoded signal transduction molecules released in multicellular organisms. The dysregulation of hormone release can cause multiple health problems and it is crucial to study these hormones for therapeutic purposes. To help the research community working in this field, we developed a prediction server that classifies hormonal peptides and non-hormonal peptides. The dataset used in this study was collected for both plants and animals from Hmrbase2 and PeptideAtlas databases. It comprises non-redundant 1174 hormonal and 1174 non-hormonal peptide sequences which were combined and divided into 80% training and 20% validation sets. We extracted a wide variety of compositional features from these sequences to develop various Machine Learning (ML) and Deep Learning (DL) models. The best performing model was logistic regression model trained on top 50 features which achieved an AUROC of 0.93. To enhance the performance of ML model, we applied Basic Local Alignment Search Tool (BLAST) to identify hormonal sequences using similarity among them, and motif search using Motif-Emerging and Classes-Identification (MERCI) to detect motifs present in hormonal and non-hormonal sequences. We combined our best performing classification model, i.e., logistic regression model with BLAST and MERCI to form a hybrid model that can predict hormonal peptide sequences accurately. The hybrid model is able to achieve an AUROC of 0.96, an accuracy of 89.79%, and an MCC of 0.8 on the validation set. This hybrid model has been incorporated on the publicly available website of HOPPred at https://webs.iiitd.edu.in/raghava/hoppred/.

show abstract

Prediction of Alzheimer’s Disease from Single Cell Transcriptomics Using Deep Learning

Srivastava,

Dhall,

Patiyal

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

Alzheimer's disease (AD) is a progressive neurological disorder characterized by brain cell death, brain atrophy, and cognitive decline. Early diagnosis of AD remains a significant challenge in effectively managing this debilitating disease. In this study, we aimed to harness the potential of single-cell transcriptomics data from 12 Alzheimer's patients and 9 normal controls (NC) to develop a predictive model for identifying AD patients. The dataset comprised gene expression profiles of 33,538 genes across 169,469 cells, with 90,713 cells belonging to AD patients and 78,783 cells belonging to NC individuals. Employing machine learning and deep learning techniques, we developed prediction models. Initially, we performed data processing to identify genes expressed in most cells. These genes were then ranked based on their ability to classify AD and NC groups. Subsequently, two sets of genes, consisting of 35 and 100 genes, respectively, were used to develop machine learning-based models. Although these models demonstrated high performance on the training dataset, their performance on the validation/independent dataset was notably poor, indicating potential overoptimization. To address this challenge, we developed a deep learning method utilizing dropout regularization technique. Our deep learning approach achieved an AUC of 0.75 and 0.84 on the validation dataset using the sets of 35 and 100 genes, respectively. Furthermore, we conducted gene ontology enrichment analysis on the selected genes to elucidate their biological roles and gain insights into the underlying mechanisms of Alzheimer's disease. While this study presents a prototype method for predicting AD using single-cell genomics data, it is important to note that the limited size of the dataset represents a major limitation. To facilitate the scientific community, we have created a website to provide with code and service. It is freely available at https://webs.iiitd.edu.in/raghava/alzscpred.

show abstract

A random forest model for predicting exosomal proteins using evolutionary information and motifs

Cited by 5 publications

References 44 publications

Prediction of anti-freezing proteins from their evolutionary profile

Prediction of anti-freezing proteins from their evolutionary profile

Prediction of peptide hormones using an ensemble of machine learning and similarity-based methods

Prediction of Alzheimer’s Disease from Single Cell Transcriptomics Using Deep Learning

Contact Info

Product

Resources

About