DAE-CFR : detecting microRNA-disease associations using deep autoencoder and combined feature representation
© 2024. The Author(s)..
BACKGROUND: MicroRNA (miRNA) has been shown to play a key role in the occurrence and progression of diseases, making uncovering miRNA-disease associations vital for disease prevention and therapy. However, traditional laboratory methods for detecting these associations are slow, strenuous, expensive, and uncertain. Although numerous advanced algorithms have emerged, it is still a challenge to develop more effective methods to explore underlying miRNA-disease associations.
RESULTS: In the study, we designed a novel approach on the basis of deep autoencoder and combined feature representation (DAE-CFR) to predict possible miRNA-disease associations. We began by creating integrated similarity matrices of miRNAs and diseases, performing a logistic function transformation, balancing positive and negative samples with k-means clustering, and constructing training samples. Then, deep autoencoder was used to extract low-dimensional feature from two kinds of feature representations for miRNAs and diseases, namely, original association information-based and similarity information-based. Next, we combined the resulting features for each miRNA-disease pair and used a logistic regression (LR) classifier to infer all unknown miRNA-disease interactions. Under five and tenfold cross-validation (CV) frameworks, DAE-CFR not only outperformed six popular algorithms and nine classifiers, but also demonstrated superior performance on an additional dataset. Furthermore, case studies on three diseases (myocardial infarction, hypertension and stroke) confirmed the validity of DAE-CFR in practice.
CONCLUSIONS: DAE-CFR achieved outstanding performance in predicting miRNA-disease associations and can provide evidence to inform biological experiments and clinical therapy.
| Medienart: |
E-Artikel |
|---|
| Erscheinungsjahr: |
2024 |
|---|---|
| Erschienen: |
2024 |
| Enthalten in: |
Zur Gesamtaufnahme - volume:25 |
|---|---|
| Enthalten in: |
BMC bioinformatics - 25(2024), 1 vom: 29. März, Seite 139 |
| Sprache: |
Englisch |
|---|
| Beteiligte Personen: |
Liu, Yanling [VerfasserIn] |
|---|
| Links: |
|---|
| Themen: |
Combined feature representation |
|---|
| Anmerkungen: |
Date Completed 01.04.2024 Date Revised 01.04.2024 published: Electronic Citation Status MEDLINE |
|---|
| doi: |
10.1186/s12859-024-05757-y |
|---|
| funding: |
|
|---|---|
| Förderinstitution / Projekttitel: |
|
| PPN (Katalog-ID): |
NLM370432258 |
|---|
| LEADER | 01000caa a22002652 4500 | ||
|---|---|---|---|
| 001 | NLM370432258 | ||
| 003 | DE-627 | ||
| 005 | 20240401233120.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 240331s2024 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.1186/s12859-024-05757-y |2 doi | |
| 028 | 5 | 2 | |a pubmed24n1359.xml |
| 035 | |a (DE-627)NLM370432258 | ||
| 035 | |a (NLM)38553698 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 100 | 1 | |a Liu, Yanling |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a DAE-CFR |b detecting microRNA-disease associations using deep autoencoder and combined feature representation |
| 264 | 1 | |c 2024 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a ƒaComputermedien |b c |2 rdamedia | ||
| 338 | |a ƒa Online-Ressource |b cr |2 rdacarrier | ||
| 500 | |a Date Completed 01.04.2024 | ||
| 500 | |a Date Revised 01.04.2024 | ||
| 500 | |a published: Electronic | ||
| 500 | |a Citation Status MEDLINE | ||
| 520 | |a © 2024. The Author(s). | ||
| 520 | |a BACKGROUND: MicroRNA (miRNA) has been shown to play a key role in the occurrence and progression of diseases, making uncovering miRNA-disease associations vital for disease prevention and therapy. However, traditional laboratory methods for detecting these associations are slow, strenuous, expensive, and uncertain. Although numerous advanced algorithms have emerged, it is still a challenge to develop more effective methods to explore underlying miRNA-disease associations | ||
| 520 | |a RESULTS: In the study, we designed a novel approach on the basis of deep autoencoder and combined feature representation (DAE-CFR) to predict possible miRNA-disease associations. We began by creating integrated similarity matrices of miRNAs and diseases, performing a logistic function transformation, balancing positive and negative samples with k-means clustering, and constructing training samples. Then, deep autoencoder was used to extract low-dimensional feature from two kinds of feature representations for miRNAs and diseases, namely, original association information-based and similarity information-based. Next, we combined the resulting features for each miRNA-disease pair and used a logistic regression (LR) classifier to infer all unknown miRNA-disease interactions. Under five and tenfold cross-validation (CV) frameworks, DAE-CFR not only outperformed six popular algorithms and nine classifiers, but also demonstrated superior performance on an additional dataset. Furthermore, case studies on three diseases (myocardial infarction, hypertension and stroke) confirmed the validity of DAE-CFR in practice | ||
| 520 | |a CONCLUSIONS: DAE-CFR achieved outstanding performance in predicting miRNA-disease associations and can provide evidence to inform biological experiments and clinical therapy | ||
| 650 | 4 | |a Journal Article | |
| 650 | 4 | |a Combined feature representation | |
| 650 | 4 | |a Deep autoencoder | |
| 650 | 4 | |a Integrated similarity | |
| 650 | 4 | |a Logistic function transformation | |
| 650 | 4 | |a MiRNA-disease associations | |
| 650 | 7 | |a MicroRNAs |2 NLM | |
| 700 | 1 | |a Zhang, Ruiyan |e verfasserin |4 aut | |
| 700 | 1 | |a Dong, Xiaojing |e verfasserin |4 aut | |
| 700 | 1 | |a Yang, Hong |e verfasserin |4 aut | |
| 700 | 1 | |a Li, Jing |e verfasserin |4 aut | |
| 700 | 1 | |a Cao, Hongyan |e verfasserin |4 aut | |
| 700 | 1 | |a Tian, Jing |e verfasserin |4 aut | |
| 700 | 1 | |a Zhang, Yanbo |e verfasserin |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t BMC bioinformatics |d 2000 |g 25(2024), 1 vom: 29. März, Seite 139 |w (DE-627)NLM109215982 |x 1471-2105 |7 nnns |
| 773 | 1 | 8 | |g volume:25 |g year:2024 |g number:1 |g day:29 |g month:03 |g pages:139 |
| 856 | 4 | 0 | |u http://dx.doi.org/10.1186/s12859-024-05757-y |3 Volltext |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a GBV_NLM | ||
| 951 | |a AR | ||
| 952 | |d 25 |j 2024 |e 1 |b 29 |c 03 |h 139 | ||