COVID-19 Patient Sera Analysis

The existing evaluation of COVID-19 patients is an empirical diagnosis method based on clinical features. When the patient has clinical features sufficient to diagnose a severe illness, the condition may have further deteriorated. Besides, due to the limited understanding of the pathogenesis of SARS-CoV-2, effective treatment of severe COVID-19 patients is still speculative.

To solve the above problems, based on the analysis data, we offer you the machine learning model for the classification of severely ill patients and provide useful clues for the diagnosis and treatment of COVID-19 patients.

Model Components

Random forest machine learning model
Our COVID-19 Patient Sera analysis system uses a random forest machine learning model based on real sample data. The model determines important variables composed of 22 protein features and 7 metabolite features and their priorities. On the training set, our model achieved an AUC(Area Under Curve) of 0.957 and an accuracy of 93.5%. In the real data validation set, our model correctly classified 23 of 29 independent patients (Figure 1).

Random forest machine learning model

Figure 1 Random forest machine learning model

Serum proteomics and metabolomics analysis model
Our Serum proteomics and metabolomics analysis model uses stable isotope-labeled proteomics strategy TMTpro and ultra-high performance liquid chromatography/tandem mass spectrometry (UPLC-MS/MS) non-targeted metabolomics methods to perform proteomics on serum samples and metabolome analysis. The model can analyze and identify 50 proteins and their 3 main pathways, namely activation of the complement system, macrophage function, and platelet degranulation (Figure 2).

Serum proteomics and metabolomics analysis model

Figure 2 Serum proteomics and metabolomics analysis model


Our COVID-19 patient sera analysis models are trained from real sample data from from 46 COVID-19 and 53 control individuals. A large amount of training data and robust algorithms make our model's accuracy rate as high as 93.5% (Area Under Curve is 0.957), which is significantly higher than manual recognition accuracy.

Supported by high-performance servers and robust algorithms, the COVID-19 patient sera analysis models can analyze a large number of sera in a short period of time and keep working 24/7, which can greatly reduce the workload of doctors and improve their diagnosis effectiveness.

Our COVID-19 patient sera analysis models can accurately and quickly complete simple screening, and make decision assistance based on AI screening results. According to the test results, our model can help junior clinic doctor to improve the quality of diagnosis, so that their diagnostic level can be improved to be equivalent to that of senior doctor in a short time.

CD ComputaBio has formed a team of experts excellent in imaging science and clinical domain knowledge, providing AI-driven solutions for COVID-19 Patient Sera Analysis according to your detailed requirements.

Online Inquiry

CD ComputaBio

Copyright © 2024 CD ComputaBio Inc. All Rights Reserved.