Identification of Protein Lysine Crotonylation Sites by a Deep Learning Framework With Convolutional Neural Networks

Abstract

Protein lysine crotonylation (Kcr) is an important type of post-translational modification that regulates various activities. The experimental approaches to identify the Kcr sites are time-consuming and it is necessary to develop computational prediction approaches. Previously, a few classifiers were based on over 100 Kcr sites from histone proteins. Recently, thousands of Kcr sites have been experimentally verified on non-histone proteins from the plant species Papaya. We found that the previous classifiers fail to identify non-histone Kcr sites. Therefore, it is necessary to develop classifiers for non-histone proteins. Accordingly, we constructed 11 different classifiers to recognize non-histone Kcr sites by combining different features and algorithms (such as random forest and convolutional neural network (CNN)). They were compared using both tenfold cross validation and independent test dataset. The classifier based on CNN and the word embedding approach, dubbed as pKcr, performed better than other classifiers. pKcr obtained AUC value of 0.855 and 0.853 for tenfold cross-validation and independent data test, respectively. No statistical difference of its performances on these two tests indicates that pKcr does not overfit. In the pKcr framework, a peptide is cleaved into biological characters followed by transformation into digital vectors. These vectors are input into the CNN with participation of multiple convolution kernels to automatically extract various features and pooling layers to perform feature selection. The superior performance of pKcr suggests that this algorithm is well suited for the Kcr prediction and may be applied broadly to predicting other types of PTM sites. pKcr can be available at http://www.bioinfogo.org/pkcr. INDEX TERMS Feature extraction, lysine crotonylation, convolutional neural network, word embedding.

DeepKhib: a deep-learning framework for lysine 2-hydroxyisobutyrylation sites prediction

“…As the DL algorithms showed superior to the traditional ML algorithms for a few PTM predictions in our previous studies [21,22] Fig. 3).…”

“…The EGAAC feature [22] is developed based on the grouped amino acids content (GAAC) feature [28,29]. In the GAAC feature, the 20 amino acid types are categorized into five groups (g1: GAVLMI, g2: FYW, g3: KRH, g4: DE and g5:…”

“…Recently, the deep learning (DL) algorithms, as the modern ML architecture, have demonstrated superior prediction performance in the field of bioinformatics, such as the prediction of modification sites on DNA, RNA and proteins [15][16][17][18][19]. We have developed a few DL approaches for the prediction of PTM sites and they all demonstrate their superiority over conventional ML algorithms [20][21][22]. Therefore, we attempted to compare the DL models with the traditional ML models for the prediction of Khib sites.…”

DeepCSO: a deep-learning network approach to predicting Cysteine S-sulphenylation sites

“…We compared the performances of these algorithms in terms of Acc, Sn, MCC, AUC and AUC01 values for both the ten-fold cross validation and the independent test (Table 1). In our previous studies DL models showed superior performance than traditional ML models [33,42]. It is still true for the CSO prediction.…”

“…LSTM WE again showed the largest AUC01 values for both ten-fold cross-validation and the independent test ( Figure 4B&D). As the encoding approach has great impact to the traditional ML models [33,42,43] and the WE approach integrated with LSTM had the best performance in this study, we attempted to investigate whether the integration of WE and RF had a good performance. Accordingly, we extracted embedding layer vector as feature encoding from LSTM WE and trained the RF model, dubbed RF WE .…”