The combination of powerful statistics and instant visualization generates exciting results. For RNA-seq data the combination of statistical filtering on expression levels with filtering on genomic entities enables focused analysis.
A wide range of statistical methods
One cornerstones is the possibility to analyze data using a flexible and easy to use statistical framework. The user can select among a wide range of statistical methods, such as:
- two group comparison (t-test)
- paired t –test
- multi group comparison (F-test) (ANOVA)
- two- way ANOVA
- linear regression
- quadratic regression
- rank regression
In addition to the methods above, filtering on Fold change and Difference are supported.
An Open API to R is integrated and opens up for use of a wide range of other statistical methods. The program is shipped with a range of tests, examples include Limma, Mann Whitney, Kruskal Wallis and Welch.
The inbuilt statistical framework (a general linear statistical model) supports the handling of eliminated factors, which means that you can remove batch effects, work with paired data or even more complex experimental set-ups.
Biomarker discovery analysis
The Biomarker Workbench is optimized for experiments and studies in the areas of drug development and biomarker discovery. Easily set-up a suite of different statistical tests to run in batch mode with the objective to select effective compounds or other relevant signals. Inspect the results in tables and directly use the structured outcome for the following analysis steps.
The Response variable option is tailored for biomarker discovery and assists in finding correlation between sample annotations – an excellent way to focus on key annotations when working with large amounts of clinical data.
Scripted work flows
Templates is the functionality used for scripting work-flows in the program. You can use built-in Templates as a quick start to standard analyses or write your own templates to simplify repetitive tasks. Templates is also the tool for integration the program into tool chains.
Survival analysis is supported with integrated statistical methods such as Hazard ratio calculations as well as visualizations in Kaplan-Meier plots.
With the NGS module and the NGS filters more options are available. Select which regions of the genome to analyze in the Genome browser, dynamically select if the regions should be restricted by read coverage, specific regions or if variants should be present or not.
Clusteringis supported in several ways; either in a semi-supervised mode using PCA or t-SNE with Projection score and variance filtering or in an unsupervised mode using kmeans++.
Classification and machine learning
The Build classifier and Classify machine learning functionality enables both the option to easy build classifiers based on models such as Boosted trees, Support Vector Machines (SVM), Random Trees (RT) and kNN and to classify new samples.
To compare and enhance your findings use the integrated GSEA Workbench which is set-up using only a few mouse-clicks. It does not get easier.
By combining the right annotations with statistical methods, data selection tools, and the eliminated factors function, a very broad range of different experiment designs can be analyzed.
Qlucore Omics Explorer also supports analysis using:
- Hierarchical clustering
- The Status panel will continuously show exactly what calculations that have been applied to your data.
Read about all features in Qlucore Omics Explorer.