Paper: ACPred-BMF: bidirectional LSTM with multiple feature representations for explainable anticancer peptide prediction.Bingqing Han, Nan Zhao, Chengshi Zeng, Xinqi Gong. Abstract: Cancer has become a major factor threatening human life and health. Under the circumstance that traditional treatment methods such as chemotherapy and radiotherapy are not highly specific and often cause severe side effects and toxicity, new treatment methods are urgently needed. Anticancer peptide drugs have low toxicity, stronger efficacy and specificity, and have emerged as a new type of cancer treatment drugs. However, experimental identification of anticancer peptides is time-consuming and expensive, and difficult to perform in a high-throughput manner. Computational identification of anticancer peptides can make up for the shortcomings of experimental identification. In this study, a deep learning-based predictor named ACPred-BMF is proposed for the prediction of anticancer peptides. This method uses the quantitative and qualitative properties of amino acids, binary profile feature to numerical representation for the peptide sequences. The Bidirectional LSTM network architecture is used in the model, and the attention mechanism is also considered. To alleviate the black-box problem of deep learning model prediction, we visualized the automatically extracted features and used the Shapley additive explanations algorithm to determine the importance of features to further understand the anticancer peptide mechanism. The results show that our method is one of the state-of-the-art anticancer peptide predictors.
Introduction:
ACPred-BMF server is used for anticancer peptide (ACP) prediction. We developed two models on different datasets:
1. [Main_model]: This model was developed on the main dataset, which used ACPs/AMPs as positive/negative samples.
2. [Alternate_model]: This model was developed on the alternate dataset, which used ACPs/random peptides as positive/negative samples.
Usage For Server:
You can click the [upload] button to submit fasta file corresponding to the peptide sequences to be predicted below, then click the [Submit] button. When your job-ID and input file display in the text box below, you can click the [Run] botton and wait for a while. When the [Click here to download] button appears below, you can proceed to download the result file. Otherwise, you may need to wait for a period of time or contact the Administrator of this website in Contact us. If you want to submit a new file, click the [Reset] button and repeat the steps above. Note that the result file is a csv file.
Try ACPred-BMF through Huggingface:
If you frequently use , we have also built a server that is easier to operate based on huggingface:
Get the Code & Dockerfile:
If you would like to download the source code, you can visit this website:
We recommend compiling ACPred-BMF(and possibly its requirements) from the source code using the latest compiler for the best performace. You can also deploy ACPred-BMF without building by Docker. You can download the Dockerfile from here
Detailed Document For ACPred-BMF:
If you would like to explore the usage details of ACPred-BMF, you can read this document:
How to read the result file:
In the results, '1' indicates that the sequence is predicted to be an anticancer peptide; '0' indicates that the sequence is predicted to be a non-anticancer peptide. The detailed interpretation of the result file is provided in the document here:
Try with Example:
We provide example files below for you to try the ACPred-BMF server. The examples contain two fasta files, where Main_test.fasta is the test set for Main_model; Alternative_test.fasta is the test set for Alternate_model.
We also provide the prediction result files on the these two datasets, where Main_prediction.csv is the prediction result on the main dataset; Alternate_prediction.csv is the prediction result on the alternate dataset.
No file has been uploaded, Please Click [Upload file] and then Click [Submit] !
