CD Genomics is able to use PLS-DA to analyze multi-omics data such as genomics, metabolomics, and proteomics, helping customers achieve high-throughput data mining.
Introduction of PLS-DA Service
Partial least squares discriminant analysis is a discriminant method in multivariate data analysis techniques and is often used to deal with classification and discrimination problems. PLS-DA employs a classical partial least squares regression model, whose response variable is a set of categorical information that responds to the class relationship between statistical units and is a supervised discriminant analysis method. The unsupervised analysis method does not discriminate among all samples, and each sample has the same contribution to the model. Therefore, unsupervised analysis methods can clearly distinguish between-group differences when the between-group differences in samples are large and within-group differences are small. However, when the between-group differences of the samples are not clear and the within-group differences are large, the unsupervised analysis method has difficulty in detecting and distinguishing between-group differences. In addition, if the differences between groups are small and the sample size of each group is different, the group with the larger sample size will dominate the model. The supervised analysis can well solve these problems encountered in the unsupervised analysis.
Figure 1. Effect of hypertension, smoking alcohol consumption and medication on metabolic profiles in sera from patients at risk for acute myocardial infarction (AMI) and sera from controls. (Khan A, et al., 2020)
Application Field
PLS-DA services can be used for but are not limited to the following research:
- Used to screen the best biomarker groups.
- Combined confocal Raman micro-spectroscopy for drug identification.
- Used to build metabolomic fingerprints.
- As a supervised identification method for plant species identification.
CD Genomics PLS-DA Service Pipeline
PLS-DA Service Content
PLS-DA is one of the most frequently used classification methods for metabolomics data analysis, which combines a regression model with dimensionality reduction and uses certain discriminant thresholds for discriminant analysis of regression results.
- We measure the strength and explanatory power of each metabolite expression pattern on the categorical discrimination of each group of samples by calculating the projected importance of variables in the PLS-DA model to help our clients screen for metabolic markers.
- Depending on the nature of the metabolites, they are ionized with different charges, respectively. For non-targeted metabolomics results, we are able to provide PLS-DA services in both positive and negative ion modes.
- Currently, the main metabolomic technology platforms are NMR, GC-MS, LC-MS, etc. We are able to provide PLS-DA services for data generated by multiple technology platforms.
We are able to use PLS-DA services to help our customers achieve high throughput data analysis in metabolomics. For questions about analysis content, project cycle, and pricing, please click online inquiry.
How It Works
CD Genomics is a professional bioinformatics service provider with years of experience in NGS and long-read sequencing (PacBio SMRT and Oxford Nanopore platforms) data analysis, integrated analysis services, database construction, and other bioinformatics solutions.
CD Genomics has professional bioinformatics experts who have successfully provided PLS-DA services to researchers in many different fields. Our professional skills and enthusiasm will provide you with high-quality analysis services. If you are interested in our services, please contact us for more details.
Reference
- Khan A, et al. High-resolution metabolomics study revealing l-homocysteine sulfinic acid, cysteic acid, and carnitine as novel biomarkers for high acute myocardial infarction risk. Metabolism. 2020 Mar; 104: 154051.
