Nemolib_App

This application uses API UWB Nemolibrary to execute the NemoProfile algorithm. It can be also used for motif detection.

We used Angular (v7.3.5) for the front end, and Spring Boot (v2.1.2.RELEASE) for the back end.
Both of the front-end and back-end applications are deployed on my AWS.
To access the back end api, please go to Swagger UI. To acces the front end web page, please go to NemoLib Application.
~~Note the above links might be invalid because I am using my personal AWS account and I only have one EC2 instance for US West.~~

To learn more, please contact wyxiao@uw.edu, hsuy717@uw.edu.

Getting Start

To run the program on your own server, please read the following

Prerequisites

Unix-like operating system (macOS or Linux), or Windows Subsystem for Linux (WSL)
Nodejs should be installed
Angular CLI should be installed
C compiler for Nauty
Java 1.8

Basic installation

Install nodejs to your system

sudo apt-get install nodejs

Install Angular

npm install -g @angular/cli

Install C and Java and all other build tools

sudo apt-get install build-essential

Using Nemolib Application

Clone the repository
Go to folder nemolib_backend and compile Spring Boot:

mvn package

Then copy the jar file under the nemolib_backend folder

cp target/<name>.jar ./

Start the spring boot application

java -jar <name>.jar

Note the server address will be localhost:8080

Go to folder nemolib_frontend and install modules:

npm install

Then start the server

ng serve

Note the server address will be localhost:4200, however you can change it to 4201 by add --port 4201
And change the permission of the labelg program: chmod u+x src/main/resources/labelg

Advanced Topics

Future Work

Although we has successfully implemented a web-based network motif application and it solves many problems that current applications have, there are still improvements we would like to work on in the future.

Modified data model

Current the frontend reads the results from the backend and process the result string to display. It is useable but there are plenty of smarter ways to do this. For example, update the output section of the Nemolibrary so the frontend can receive more modularized data.

Add file management for back-end

Currently we are using Linux server for our back-end Spring application. All uploaded files were saved in side the server and have to be deleted manually. There should be a new API for the Storage service to control those files.

Add feature to read direct graph

The current application can only detect network motif based on the an undirected graph. We can improve it by implementing another class which can read direct graph and detect network motif.

Provide motif visualization

Our web application can provide user the motif data including graph label, relative frequency, random mean frequency, Z score, P value, and NemoProfile . However, the data is in text format. Since network is essentially graph, if we can add the visualization for each motif and where they are in a network, it will give users more insights about the results.

lytbfml/Nemolib_App