Advance Data Analysis

For advance analysis we work closely with our clients and, depending upon the requirement, advance analysis is performed for biological/ clinical decision making. Our advance analysis services include:

• Principal Component Analysis
• Post Hoc Test
• Clustering
• Biological Pathway Analysis

Principal Component Analysis:

We use Principal Component analysis to compress the variability of gene expression profiles into small number of  “principal component” (usually involving tens of thousands of genes to 2 or 3) so that differences among various conditions/ treatments can be visualized in a 2 or 3 dimensional plot.

Post Hoc Test:

Applied only to preselected differentially expressed genes based on ANOVA. Tukey and Scheff_e (student t- test) methods are used to compare variables for differentially expressed genes. The Bonferroni method is used to conservatively reduce errors due to testing of multiple templates.

Clustering:

SImilarities in gene expression profiles across the array data within an experiment are detected by clustering. The expression profiles of gene describe how the relative expression level changes over time course or under the influence of various / treatments. Genes with similar profiles are grouped together. Unknown novel genes can be identified and clustered with known genes. Clustering enables a new scale of biological experimentation, open doors to higher-confidence target discovery, time series, disease classification and pathway analysis, regulatory and functional group analysis. Advanced Genomics uses various clustering methods to meet the goals of the clients. The most commonly used clustering techniques are hierarchical clustering and K-mean clustering.

Hierarchical clustering: genes are ordered in a dendrogram according to the degree of similarity between their expression profiles across a set of conditions. Genes can be selected within Heatmap and relative expression level histogram illustrates relative expression level. Gene Ontology information can be easily obtained by passing cursor over any node of gene tree.

K-mean clustering: Genes are divided into several groups and genes in the same group have similar gene expression pattern, while genes in different groups have different gene expression  pattern. The gene expression  pattern and number of genes in each group are visualized in histogram and line graph thumbnail view.

Biological Pathway Analysis:

Valuable data is qualified and quantified with biological pathway analysis. Some selected genes may be involved in some well established Pathways. By comparing gene expression levels of genes in the pathway collectively, we can find out whether a particular pathway is significantly affected by particular treatment. Advanced Genomics has the full range of softwares set up to get the maximum information about biological pathways from analyzed gene expression data. Expression analysis systematic explorer (EASE) software is used for gene ontology and annotation. Softwares like Metacore, Kyoto encyclopedia of genes and genomes (KEGG), EASE, GeneSight and GeneSifter, GENMAPP, BBID, GeneGo are used to relate data with biological pathways. Genes related to various pathways are selected manually for expression analysis among various conditions.

Discover Potential Biomarkers: Identified differentially expressed genes are compared within various experimental parameters and highly expressed and regulated genes that are related with known pathways, diseases or regulatory  functions are selected as potential biomarkers.