This project consists of a backend and a frontend project namely omics-api and omics-fe. Omics API is a NestJs application and FE is a react app.
These projects are independent so you can run them independently. Please check the projects' details for installation and run.
In the application UI, you have two links, "Gene Visualization" and "Create Experiment" in the header. To create dummy data, please enter the page "Create Experiment".
On this page, you must fill Experiment name, number of samples, Number of Genes, and Outlier ratio fields. When you click on "Create mock data" the backend service will generate mock data based on the parameters you provided.
The mock data will generate expression values between 120 and 150. However, to generate some outliers, with the probability of outlearRatio, it'll generate expression values between 1 and 190.
The names of the genes will be like "Gene{a number btw 1000 and 51000}"
Sample names will be like "Sample{a number btw 0 and # of samples}"
On the visualization page, you need to select an experiment. Then you have two options:
There is a simple service that calculates outliers based on a z-score threshold. You can set a z-score threshold and click on the Get Outliers button. Once you click, you will see a table showing genes with outliers and samples having outliers.
This functionality is not designed for handling big data. We can use some ML algorithms and an asynchronous mechanism to find better outliers.
Under the Outliers table, you can visualize gene expression data. It would be best if you typed some text to search genes. Because of the naming of the genes in the mock data, you can enter some numbers and select more than one gene in the dropdowns.
After selecting genes, you need to click on the "Get Gene Data" button. This will get the expression values for the genes. The system will draw a heatmap for the selected genes, too.
Once you click on the analyze button for the genes in the table, you will see the statistics and a scatter plot for expression values.
If you want to observe the structure of the data please check the dummy data that I created.