Team:NCTU Formosa/Demonstrate

navigation

NCTU_Formosa: Demonstration
Parabase Website
- the demonstration of Peptide Prediction

     To present the work we did, we established a website named Parabase to demonstrate our peptide prediction system and the search system, showing their good functioning and convenience.

Content:

  1. The antifungal peptide scoring system
  2. The search system

The website provides two functions:

1. Antifungal Prediction System:

     With SCM, we have already done the prediction of several peptides obtained from UniProt. Peptides' probability to be antifungal are shown on our website. Users can also do the prediction by themselves. Online prediction system is available on the website, which supports fasta format inputs of both manuals inputting amino acid sequences and files uploading. The output of the prediction of a peptide will be a score presenting the probability of having antifungal activity.

Figure 1: Antifungal Scoring Card

2. The Search System:

     We have collected hosts, pathogens, and antifungal peptides. The data stored in the database can be obtained via tagging keywords in the searching engine on the website. The keyword can be the scientific name for hosts and pathogens, and for searching antifungal peptides, the name or the amino acid sequence of the peptide is available.

     The searching results come out with the relationship between hosts, pathogens, and peptides. For searching the host, the pathogens that infect the one will be presented, while the pathogen to present the hosts and the peptides that inhibit the growth of the pathogen. And for searching the peptide, amino acid sequences, pathogens, sources, abstracts, and links to the references of the peptide will be given.

Figure 2: The main page of the Parabase website

     By our website, users can quickly find out the information such as: Which pathogens are also affected by the peptide? Which peptides may be helpful to a particular crop. It helps users to find peptides that are useful for preventing crop infection by a particular pathogen while not killing other pathogens that are not harmful to the crop.

Front End:

     For the front end, we used CSS to prettify the layout of the web page.

Back End:

     On the back end, we gained our web server through NCTU Web Hosting service. We used PHP and Python to respond to web requests such as searching hosts, pathogens, antifungal peptides and doing antifungal peptide prediction. We also used SQLite for data storage.

Plant Box Demonstration
- The demonstration of the Disease Occurrence Prediction

     To present the work we achieved, we made a plant box to simulate our Disease Occurrence System.

Content:

  1. Demonstration video
  2. Data sensing
  3. Prediction model
  4. Spraying system
  5. Data showing

     To present the work we achieved, we made a plant box to simulate our Disease Occurrence system.

Video 1: IoT demonstration

     We divided this system into four parts, including data sensing, prediction model, spraying system,and data showing.

1. Data sensing:

     The main purpose is to detect real-time environmental data such as relative humidity, temperature, rainfall, and pressure. Via the operation of the sensor, we can collect these data immediately. Then through the transmission of the Internet, we can send these data to IoTtalk for data integration.

Sensors:

     - Humidity and temperature sensor: DHT22

     The DHT22 helps to measure the temperature and humidity. Through using it we can get the daily weather situation and receive the necessary information.

     - Atmospheric pressure sensor: BMP180

     BMP180 is mainly to measure the instantaneous atmospheric pressure, and it is relative to the altitude of the sea level.

     - Distance sensor: HC-SR04

     The initial function of the HC-SR04 is to measure the short distance. Depending on this feature, we wrote the code to convert it into rainfall sensor.

2. Prediction model:

     This part is the core of the whole system. We have set up another server to read the data stored in IoTtalk, then the model would calculate the probability of the diseases happening, and passed back to IoTtalk.

(For more detail, please check out Disease Occurrence Prediction & Disease Occurrence Model)

3. Spraying system:

     This part is used to judge the probability of the disease occurrence prediction model. First of all, we would assume a threshold. If the calculating result was higher than the threshold, spraying system would be turned on, and sprayed the antifungal biopesticides.

4. Data showing:

     Because of the assistance of the ThingSpeak website, patterns you wanted to observe could be shown, such as bar graphs, line graphs…etc. Thus, data collected on the IoTtalk platform would be shown in a convenient way.

Github

     Parabase Website

     IoT

     Peptide Prediction Model

Untitled Document

×

Loading ...