ProbeSearchV3

Overview

Probesaerch align primer/probe/gRNA sequences against selected genome databases. It currently has two versions: V3 and V3.2. The app is written in Nodejs, Expressjs, and Vue.js.

Installation

1. Install Node.js: https://nodejs.dev/learn/how-to-install-nodejs
2. Install Razers3: Follow the instructions provided in the SeqAn GitHub repository to set up Razers3 aligner executable.
3. setup the app

git clone https://github.com/TsailabBioinformatics/ProbeSearchV3
cd ProbeSearchV3
npm install

cd vue
npm install

# Update the directories path of `data/` and `indices/` in the script. Refer to [Editing Backend section.](#editting-probesearchv3)
node app.js

Deployment

pm2 is a node process manager that keeps the node app runnning as a background process. Run pm2 list to checkout the current node processes running on pm2. If for some reason probesearch or probesearchv3.2 is no longer running, we can run pm2 start app.js --name probesearch to restart it. Also, one can run pm2 --help to learn about all of its functions.

Understanding ProbeSearchV3

Here is description of full stack of ProbeSearchV3.

Directory Structure

• data/: Fasta files are located in /data/probesearchDB/data/ on the server
• bowtie/: Bowtie2 executables are located in bowtie/
• scripts/: Helper scripts for Bowtie2
• indices/: Bowtie2 uses indexed versions of the fasta files. Indices are located in indices/.
  • In order to create an index for bowtie, use ./bowtie/bowtie2-build <path_to_fasta> <index_name>. Then move these indices to indices/
  • In order to create an index for batmis, use ./usr/local/bin/build_index <path_to_fasta> <index_name>. Then move these indices to indices/
  • Indices are integral to bowtie2 alignment. app.js makes a call to ./bowtie/bowtie2 -x indices/<db> -k 30 -c <read>. Here, <db> represents the user-selected database, and <read> represents the input read.
• GeneTable/: contains gene database used by V3.2
• vue/: Frontend Vue files

Data Flow
The following details how data moves in the app. We'll see how the client-side form submission request leads to the eventual alignment result response.

1. Initial Input Processing
   User input is first handled by SequenceForm.vue. Within SequenceForm.vue, these data are tracked as read: '' and db: []. read is a string because it represents a primer/probe/gRNA sequence, and db is an array because it represents a set of genome databases. When the user fills out the form, the variables are defined respetively. Then, when the user submits the form (e.g., clicks Search), the variables are propagated, or emitted (in Vue lingo), up to its parent component App.vue. App.vue, then processes the user input and makes a put request to the Express api, app.js (axios.put('/', payload)).

2. API Functions
   The put function first calls a child process: ./bowtie/bowtie2 -x indices/ + String(req.body.db) -k 30 -c + String(req.body.read) + --end-to-end --no-hd. Note, req.body.db and req.body.read is the Express way of accessing the inputted db and read respectively. This child process initiates the bowtie2 alignment function, which returns a SAM file as stdout. Then, the SAM file, or stdout is parsed and molded into an alignment illustration. The illustration of the alignmened is thne returned to App.vue.

3. Returning Alignment Result
   app.js sends back data in the form of an illustration to App.vue. In turn, App.vue passes this data to a child component called AlignmentResult.vue. Once the AlignmentResult component(s) receive the data, they show on client side.

UI

Both versions share the same user interface to avoid complexity. However, their backend implementations differ.

Scripts

Each version has three separate scripts:

1. Bowtie2 Scripts:

   • Handles alignment using Bowtie2 for all genome databases except sPta717V2.0.
   • V3: app.js
   • V3.2: appV3.js

2. Specialized Bowtie2 Script:

   • Specifically for the sPta717V2.0 database due to differences in output format.
   • Script: 717V2.js

3. Razers3 Scripts:

   • Utilizes Razers3 aligner for alignment.
   • V3: razers3.js
   • V3.2: razers3V3.js

Entry Points

• V3: Entry point is through app.js.
• V3.2: Entry point is through appV3.js.

Editting ProbeSearchV3

• Frontend
  • The frontend can be editted from within the vue/ directory.
  • The three components as of now are the parent, App.vue, and its children, SequenceForm.vue & AlignmentResult.vue.
  • After editting the vue files, run npm run build from within the vue/ directory. This command builds the static HTML, css, and js for the frontend and places them in vue/dist/ directory. The express app, app.js, references these static files automatically, so that's all: edit then build.
• API
  • The main app can be editted at app.js or appV3.js based on the version of the app. It is written in Express.
• Backend
  • The backend consists of two mains parts: the fasta files & the indexed fasta files. In order to edit the backend, or implement more genomes, do the following:
  1. Transfer whatever fasta files you wish to the parent directory.
  2. Build the indices: ./bowtie/bowtie2-build <path_to_fasta> <index_name>. Name the index (second parameter) the same name as you display the database on the frontend.
  3. Move the raw fasta file to the data/ directory, and move the indexed fasta files to the indices/ directory.
  4. Go the Express app.js, and add to db_dictionary, which facilitates a dictionary between the frontend name of the database and the location of the fasta file.

Versions

1. V3:

   • Allows users to input primer/probe/gRNA sequences and select a set of genome databases.
   • Upon submission, the app returns the result of the input read aligned against the selected genome using the Bowtie2 software.
   • Provides an option to change the aligner to Razers3, primarily used for aligning with short reads.

2. V3.2:

   • Includes all features of V3.
   • Additionally provides Gene Features and CDS Overlapping Info.

Note:

Make sure all the files - app.js, appV3.js, razers3.js, razers3V3.js and 717V2.js have update path according to respective directories.