Browse Source

更新 'README.md'

master
Katherine1216 2 weeks ago
parent
commit
7c46f14a9b
1 changed files with 25 additions and 31 deletions
  1. +25
    -31
      README.md

+ 25
- 31
README.md View File

@@ -1,43 +1,37 @@
# **Bringing Structural Implications and Deep Learning Based Drug Prediction for KRAS Mutants**
# DTI-CDF
## An `instruction` file for DTI-CDF for DTIs prediction.

## Aamir Mehmood, Aman Chandra Kaushik, Cheng-Dong Li, Dong-Qing Wei
### Introduction
Drug-target interactions (DTIs) play a crucial role in target-based drug discovery and development. Computational prediction of DTIs has become a popular supplementary strategy to conduct the experimental methods, which are both time as well as resource consuming, for identification of DTIs. However, the performances of the current DTIs prediction approaches suffer from a problem of low precision and high false positive rate. In this study, we aim to develop a novel DTIs prediction method, named DTI-CDF, for improving the prediction performance based on a cascade deep forest model with multiple similarity-based features extracted from the heterogeneous graph. In the experiments, we build five replicates of 10-fold cross-validations under three different experimental settings of data sets, namely, corresponding DTIs values of certain drugs (SD), targets (ST), or drug-target pairs (SP) in the training sets are missed but existed in the test sets. The experimental results demonstrate that our proposed approach DTI-CDF achieves significantly higher performance than of the state-of-the-art methods. And there are 1352 predicted new DTIs are proved correct by KEGG and DrugBank databases.

College of Life Sciences and Biotechnology, The State Key Laboratory of Microbial Metabolism, Shanghai Jiao Tong University;
Wuxi School of Medicine, Jiangnan University, Wuxi, China

#### Abstract
The Colorectal cancer is considered as one of the leading causes of death that is linked with the KRAS harboring codon 13 and
codon 61 mutations. The drive of the current work is to search for clinically important codon 61 mutations and analyze how they affect the
protein structural dynamics. Additionally, Deep-learning and neural-network approaches are practiced to scan for potential compounds that
might have a comparatively better affinity. Public databases like Cancer Genome Atlas (TCGA) and Genomic Data Commons (GDC) were accessed
for obtaining the data regarding mutations that are associated with colon cancer. Multiple analysis such as genomic alteration landscape,
survival analysis, and systems-biology based kinetic simulations were carried out to predict dynamic changes for the chosen mutations.
A molecular dynamics simulation of 100 ns for all the seven shortlisted codon 61 mutations have been conducted that revealed noticeable
deviations. Finally, the predicted compounds via a machine learning-based approach were docked with the KRAS 3D conformer and it was
observed that the proposed compounds have better affinities and docking score as compared to the already existing recommended drugs.
Taking together the outcomes of systems biology and molecular dynamics, it is observed that the reported mutations in SII region are
highly deleterious as they have an immense impact the protein sensitive sites native conformation and overall stability. The drugs
reported in this study also has significant performance and are encouraged to be used for further evaluation regarding the situation
that ascends as a result of KRAS mutations.
### Requirements
This method developed with Python 2.7, please make sure all the dependencies are installed, which is specified in DTI-CDF_requirements.txt.


#### Data
The background data for the Systems Biology and Machine Learning Scripts are provided in separate folders.
The integrase.cv is a huge file and is thus made accessible at https://drive.google.com/open?id=1-lwMZqWrLVS9UNae7Qoe9VGMJ0HXpZis.
In case of any inconvinience, please feel free to reach us at iaamirofficial@gmail.com.
### Reference
DTI-CDF a cascade deep forest model towards the prediction of drug-target interactions based on hybrid features.

#### Demo
#### Instruction to execute this Model
Run data set (as a demo)
Download integrase.csv file (availabe at https://drive.google.com/open?id=1-lwMZqWrLVS9UNae7Qoe9VGMJ0HXpZis)
In the Machine Learning Data folder,
Run the DML Training script (Python).py
Open CMD and input:
cd .\Machine Learning Data\Your_Folder
python -u Script.py > Example.out

### Run NR data set (as a demo)
1. Download fold “.\DTI-CDF\2_Example_NR”.

2. In the “.\DTI-CDF\2_Example_NR” path, run the Example_NR.py file, as follows:
Open CMD and input:
`cd .\DTI-CDF\2_Example_NR`
`python -u Example_NR.py > Example_NR.out`


Please see “Example_NR.out” file for the results/outputs which contains the results of performance metrics, time required for the program to run and the new DTIs predicted by this method.
If you want to try other data sets, just follow this demo, and the codes and data have been supported in fold “1_all_code” and “1_original_data”, respectively.

### Package dependencies

The package is developed in python 2.7, higher version of python is not suggested for the current version.
Run the following command to install dependencies before running the code: pip install -r DTI-CDF_requirements.txt.
If something wrong when you run the code, you could reinstall gcforest as follow: move fold "DTI-CDF/gcforest" to your python environment, such as dir = '\Anaconda3\envs\ipykernel_py2\Lib\site-packages'.

### Others
Please read reference and py file for a detailed walk-through.

### Thanks

Loading…
Cancel
Save