This is a monorepo containing the front-end and back-end of the Exscientia code challenge (exscientia-fe and exscientia-be directories, respectively). Details of the decisions made, technologies used, and potential improvements can be found in the respective README.md files for both the front-end and back-end (exscientia-fe/README.md and exscientia-be/README.md).
Using the provided data, I have created a full-stack application which allows users to view molecular structures based on their SMILES representation.
The front-end is a React application which makes use of a third-party package to parse SMILES strings and draw their molecular structure. There are two main features of the front-end which make use of this functionality:
- SMILES Parser: Users can input their own SMILES strings into a textbox and see the appropriate molecule drawn in real-time.
- Server Molecules: On page load, the application will send a GET request to the back-end for the compounds given in the brief (these have been seeded into the database which the back-end interacts with). It will then display these molecules, along with their molecular weight and molecular formula, as cards in a grid format.
The front-end is somewhat responsive, as the above grid-layout will adjust to screen size. However, due to a limitation of the SMILES parsing package, the site would need much more work to be truly responsive (e.g. makes heavy use of <canvas> with specific pixel counts).
The back-end is comprised of a Go HTTP server and a Mongo database. These are both dockerized, and can be brought up with a single docker-compose up command (see exscientia-be/README.md for details). The database will be automatically seeded on start-up with the compounds.json data which was provided, using a seeding script.
This application has a few prerequisites:
- npm
- node
- docker
Once these are installed, bring up the application as follows:
cdinto theexscientia-feandexscientia-bedirectories, and follow the instructions laid out in the respectiveREADME.mdfiles to bring up each part of the application.- Note, both the front-end and back-end must be brought up for full functionality of the application.
As this challenge was somewhat time-bound, there are a few areas which would need particular attention and some improvements which should be made if this was to be a "real" application. I have detailed these in the respective README.md files for the front/back-end, however the lists are not exhaustive.
I have also made heavy use of comments in code to explain decisions or suggest alterations I would have made if this was an application to be maintained for a long time.
Below is the brief given for this challenge.
You’ve been given a data set as a json file with 100 records in it, the challenge is to showcase some interesting software engineering with the data set.
The data set is all about compounds and their assay results which represent a much simplified view of the data we deal with in the drug discovery domain. Don’t worry if you don’t have a clue about drug discovery, it’s all just pictures, strings and numbers the same as every other business. Each compound can have one or more assay result so it’s just a standard master detail scenario.
There are two files, compounds.json is the data file and schema.json describes the data (fields, types and so on). There is also a subfolder called images which contains 100 png images, one for each record in the data set.
- Don’t incur any costs, free tools are available.
- Use whatever tech stack you prefer that will show off your skills to the max.
- You have up to 1 week to complete the challenge, we appreciate you are busy with life and work so if you need longer let us know early on or consider doing something simple.
- Please check your solution into github (or similar) and share the repository with REDACTED EMAIL.
- The repository must contain a README file (.md is good) that describes the very simple and easy to follow steps needed to use / start up your solution.
- Include in your repo any build / deploy scripts, we are interested in your whole solution.
If you are a backend sort of person why not create a script that loads the compounds.json data into a database then expose the data via an API.
If you are a front end sort of a person why not render compounds.json in a master detail web app, there are enough fields in the file to create a gitlab runners simple chart too.