TransGeneSelector: A Transformer-based Approach Tailored for Key Gene Mining with Small Plant Transcriptomic Datasets DOI Creative Commons
Kerui Huang,

Jianhong Tian,

Лэй Сун

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: Sept. 28, 2023

Abstract Gene mining, particularly from small sample sizes such as in plants, remains a challenge life sciences. Traditional methods often omit significant genes, while deep learning techniques are hindered by constraints and lack specialized gene mining approaches. This paper presents TransGeneSelector, the first method tailored for key transcriptomic datasets, ingeniously integrating data augmentation, filtering, Transformer-based classifier. Tested on Arabidopsis thaliana seeds’ germination classification using just 79 samples, it not only achieves performance par with, if superior to, Random Forest SVM but also excels identifying upstream regulatory genes that might miss, these pinpointed more accurately reflect metabolic processes inherent seed germination. TransGeneSelector’s ability to mine vital limited datasets signifies its potential current state-of-the-art scenarios, providing an efficient versatile solution this critical research area.

Language: Английский

A Review on Scan Strategies in Laser-based Metal Additive Manufacturing DOI Creative Commons
M. Junaid Dar,

Andre Georges Ponsot,

Caden Jacob Jolma

et al.

Journal of Materials Research and Technology, Journal Year: 2025, Volume and Issue: unknown

Published: April 1, 2025

Language: Английский

Citations

0
Optimal gene therapy network: Enhancing cancer classification through advanced AI-driven gene expression analysis DOI Creative Commons

Tulasi Raju Nethala,

Bidush Kumar Sahoo,

P. Srinivasulu

et al.

e-Prime - Advances in Electrical Engineering Electronics and Energy, Journal Year: 2024, Volume and Issue: 7, P. 100449 - 100449

Published: Feb. 14, 2024

Gene therapy is an advanced medical approach that aims to find solutions for various cancers by identifying optimal gene expressions. In this context, computer-aided detection of expressions becomes a research challenge, where artificial intelligence methods are employed classify cancer types. However, traditional machine learning models must be improved accurately classifying cancers, leading unsatisfactory quantitative performance. Therefore, work implemented the network (OGT-Net) different types from expression sequences. Initially, dataset pre-processing operation normalizes dataset, which maintains uniform nature all records in dataset. Then, light gradient boosting model (LGBM) extracts correlated features pre-processed contains relationship among data. addition, interrupt-based Harris Hawk optimization (IHHO) LGBM data, decreasing total number removing redundant customized deep convolution neural (DLCNN) used categorize diseases using datasets based on lymphography, colon, lung, ovarian, and prostate cancers. The simulation results reveal proposed OGT-Net performance compared existing approaches, with average accuracy 91.128%, precision 90.836%, recall 91.25%, F1-score 90.7%.

Language: Английский

Citations

2
Segmentation for mammography classification utilizing deep convolutional neural network DOI Creative Commons
Dip Kumar Saha,

Tuhin Hossain,

Mejdl Safran

et al.

BMC Medical Imaging, Journal Year: 2024, Volume and Issue: 24(1)

Published: Dec. 18, 2024

Mammography for the diagnosis of early breast cancer (BC) relies heavily on identification masses. However, in stages, it might be challenging to ascertain whether a mass is benign or malignant. Consequently, many deep learning (DL)-based computer-aided (CAD) approaches BC classification have been developed. Recently, transformer model has emerged as method overcoming constraints convolutional neural networks (CNN). Thus, our primary goal was determine how well an improved could distinguish between and malignant tissues. In this instance, we drew Mendeley data repository's INbreast dataset, which includes types. Additionally, segmentation anything (SAM) used generate optimized cutoff region interest (ROI) extraction from all mammograms. We implemented successful architecture modification at bottom layer pyramid (PTr) identify mammography images. The proposed PTr using transfer (TL) approach with technique achieved best accuracy 99.96% binary classifications area under curve (AUC) score 99.98%, respectively. also compared performance other vision transformers (ViT) DL models, MobileNetV3 EfficientNetB7, study, modified prediction image approaches. Data techniques accurately regions affected by BC. Finally, classified tissues, vital radiologists guide future treatment.

Language: Английский

Citations

2
Survival Analysis for Multimode Ablation Using Self-Adapted Deep Learning Network Based on Multisource Features DOI Open Access
Ziqi Zhao, Wentao Li,

Ping Liu

et al.

IEEE Journal of Biomedical and Health Informatics, Journal Year: 2023, Volume and Issue: 28(1), P. 19 - 30

Published: March 31, 2023

Novel multimode thermal therapy by freezing before radio-frequency heating has achieved a desirable therapeutic effect in liver cancer. Compared with surgical resection, ablation treatment relatively high risk of tumor recurrence. To monitor progression after ablation, we developed novel survival analysis framework for prediction and efficacy assessment. We extracted preoperative postoperative MRI radiomics features vision transformer-based deep learning features. also combined the immune from peripheral blood responses using flow cytometry routine tests treatment. selected random forest improved Cox mixture (DCM) analysis. properly accommodate multitype input features, proposed self-adapted fully connected layer locally globally representing evaluated method our clinical dataset. Of note, rank highest feature importance contribute significantly to accuracy. The results showed promising C td-index 0.885 ±0.040 an integrated Brier score 0.041 ±0.014, which outperformed state-of-the-art combinations prediction. For each patient, individual probability was accurately predicted over time, provided clinicians trustable prognosis suggestions.

Language: Английский

Citations

5
AI-driven transcriptomic encoders: From explainable models to accurate, sample-independent cancer diagnostics DOI Creative Commons
Danilo Croce, Artem Smirnov, Luigi Tiburzi

et al.

Expert Systems with Applications, Journal Year: 2024, Volume and Issue: 258, P. 125126 - 125126

Published: Aug. 23, 2024

Citations

0
PlasGO: enhancing GO-based function prediction for plasmid-encoded proteins based on genetic structure DOI Creative Commons
Yongxin Ji, Jiayu Shang, Jiaojiao Guan

et al.

GigaScience, Journal Year: 2024, Volume and Issue: 13

Published: Jan. 1, 2024

Plasmid, as a mobile genetic element, plays pivotal role in facilitating the transfer of traits, such antimicrobial resistance, among bacterial community. Annotating plasmid-encoded proteins with widely used Gene Ontology (GO) vocabulary is fundamental step various tasks, including plasmid mobility classification. However, GO prediction for faces 2 major challenges: high diversity functions and limited availability high-quality annotations.

Language: Английский

Citations

0
Federated learning for oncology: Breast Cancer subtyping Case Study Using Gene Expression DOI

Karl Paygambar,

Mallek Mziou-Sallami,

Fabien Baligand

et al.

2021 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), Journal Year: 2024, Volume and Issue: unknown, P. 4474 - 4480

Published: Dec. 3, 2024

Language: Английский

Citations

0
Identification of Key Features in Breast Cancer Diagnosis Using Vision Transformer DOI
Jingyi Dong,

Wanqiong Huang,

J.-J. Zhang

et al.

Published: Oct. 18, 2024

Language: Английский

Citations

0
TransGeneSelector: A Transformer-based Approach Tailored for Key Gene Mining with Small Plant Transcriptomic Datasets DOI Creative Commons
Kerui Huang,

Jianhong Tian,

Лэй Сун

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: Sept. 28, 2023

Abstract Gene mining, particularly from small sample sizes such as in plants, remains a challenge life sciences. Traditional methods often omit significant genes, while deep learning techniques are hindered by constraints and lack specialized gene mining approaches. This paper presents TransGeneSelector, the first method tailored for key transcriptomic datasets, ingeniously integrating data augmentation, filtering, Transformer-based classifier. Tested on Arabidopsis thaliana seeds’ germination classification using just 79 samples, it not only achieves performance par with, if superior to, Random Forest SVM but also excels identifying upstream regulatory genes that might miss, these pinpointed more accurately reflect metabolic processes inherent seed germination. TransGeneSelector’s ability to mine vital limited datasets signifies its potential current state-of-the-art scenarios, providing an efficient versatile solution this critical research area.

Language: Английский

Citations

0