Anomaly Detection Models for SARS-CoV-2 Surveillance Based on Genome <i>k</i>-mers
Since COVID-19 has brought great challenges to global public health governance, developing methods that track the evolution of the virus over the course of an epidemic or pandemic is useful for public health. This paper uses anomaly detection models to analyze SARS-CoV-2 virus genome <i>k</...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-11-01
|
| Series: | Microorganisms |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-2607/11/11/2773 |
| _version_ | 1827639176033468416 |
|---|---|
| author | Haotian Ren Yixue Li Tao Huang |
| author_facet | Haotian Ren Yixue Li Tao Huang |
| author_sort | Haotian Ren |
| collection | DOAJ |
| description | Since COVID-19 has brought great challenges to global public health governance, developing methods that track the evolution of the virus over the course of an epidemic or pandemic is useful for public health. This paper uses anomaly detection models to analyze SARS-CoV-2 virus genome <i>k</i>-mers to predict possible new critical variants in the collected samples. We used the sample data from Argentina, China and Portugal obtained from the Global Initiative on Sharing All Influenza Data (GISAID) to conduct multiple rounds of evaluation on several anomaly detection models, to verify the feasibility of this virus early warning and surveillance idea and find appropriate anomaly detection models for actual epidemic surveillance. Through multiple rounds of model testing, we found that the LUNAR (learnable unified neighborhood-based anomaly ranking) and LUNAR+LUNAR stacking model performed well in new critical variants detection. The results of simulated dynamic detection validate the feasibility of this approach, which can help efficiently monitor samples in local areas. |
| first_indexed | 2024-03-09T16:34:55Z |
| format | Article |
| id | doaj.art-11d412dd1aed4e4f9d0eb04dbde92edd |
| institution | Directory Open Access Journal |
| issn | 2076-2607 |
| language | English |
| last_indexed | 2024-03-09T16:34:55Z |
| publishDate | 2023-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Microorganisms |
| spelling | doaj.art-11d412dd1aed4e4f9d0eb04dbde92edd2023-11-24T14:57:13ZengMDPI AGMicroorganisms2076-26072023-11-011111277310.3390/microorganisms11112773Anomaly Detection Models for SARS-CoV-2 Surveillance Based on Genome <i>k</i>-mersHaotian Ren0Yixue Li1Tao Huang2Bio-Med Big Data Center, CAS Key Laboratory of Computational Biology, Shanghai Institute of Nutrition and Health, University of Chinese Academy of Sciences, Chinese Academy of Sciences, Shanghai 200031, ChinaBio-Med Big Data Center, CAS Key Laboratory of Computational Biology, Shanghai Institute of Nutrition and Health, University of Chinese Academy of Sciences, Chinese Academy of Sciences, Shanghai 200031, ChinaBio-Med Big Data Center, CAS Key Laboratory of Computational Biology, Shanghai Institute of Nutrition and Health, University of Chinese Academy of Sciences, Chinese Academy of Sciences, Shanghai 200031, ChinaSince COVID-19 has brought great challenges to global public health governance, developing methods that track the evolution of the virus over the course of an epidemic or pandemic is useful for public health. This paper uses anomaly detection models to analyze SARS-CoV-2 virus genome <i>k</i>-mers to predict possible new critical variants in the collected samples. We used the sample data from Argentina, China and Portugal obtained from the Global Initiative on Sharing All Influenza Data (GISAID) to conduct multiple rounds of evaluation on several anomaly detection models, to verify the feasibility of this virus early warning and surveillance idea and find appropriate anomaly detection models for actual epidemic surveillance. Through multiple rounds of model testing, we found that the LUNAR (learnable unified neighborhood-based anomaly ranking) and LUNAR+LUNAR stacking model performed well in new critical variants detection. The results of simulated dynamic detection validate the feasibility of this approach, which can help efficiently monitor samples in local areas.https://www.mdpi.com/2076-2607/11/11/2773anomaly detectionvirus surveillanceSARS-CoV-2<i>k</i>-mermachine learning |
| spellingShingle | Haotian Ren Yixue Li Tao Huang Anomaly Detection Models for SARS-CoV-2 Surveillance Based on Genome <i>k</i>-mers Microorganisms anomaly detection virus surveillance SARS-CoV-2 <i>k</i>-mer machine learning |
| title | Anomaly Detection Models for SARS-CoV-2 Surveillance Based on Genome <i>k</i>-mers |
| title_full | Anomaly Detection Models for SARS-CoV-2 Surveillance Based on Genome <i>k</i>-mers |
| title_fullStr | Anomaly Detection Models for SARS-CoV-2 Surveillance Based on Genome <i>k</i>-mers |
| title_full_unstemmed | Anomaly Detection Models for SARS-CoV-2 Surveillance Based on Genome <i>k</i>-mers |
| title_short | Anomaly Detection Models for SARS-CoV-2 Surveillance Based on Genome <i>k</i>-mers |
| title_sort | anomaly detection models for sars cov 2 surveillance based on genome i k i mers |
| topic | anomaly detection virus surveillance SARS-CoV-2 <i>k</i>-mer machine learning |
| url | https://www.mdpi.com/2076-2607/11/11/2773 |
| work_keys_str_mv | AT haotianren anomalydetectionmodelsforsarscov2surveillancebasedongenomeikimers AT yixueli anomalydetectionmodelsforsarscov2surveillancebasedongenomeikimers AT taohuang anomalydetectionmodelsforsarscov2surveillancebasedongenomeikimers |