Gain of function mutations in fibroblast growth factor (FGF) receptors cause chondrodysplasia and craniosynostosis syndromes. The ligands interacting with FGF receptors (FGFRs) in developing bone have remained elusive, and the mechanisms by which FGF signaling regulates endochondral, periosteal, and intramembranous bone growth are not known. Here we show thatFgf18is expressed in the perichondrium and that mice homozygous for a targeted disruption ofFgf18exhibit a growth plate phenotype similar to that observed in mice lackingFgfr3and an ossification defect at sites that expressFgfr2. Mice lacking eitherFgf18orFgfr3exhibited expanded zones of proliferating and hypertrophic chondrocytes and increased chondrocyte proliferation, differentiation, and Indian hedgehog signaling. These data suggest that FGF18 acts as a physiological ligand for FGFR3. In addition, mice lackingFgf18display delayed ossification and decreased expression of osteogenic markers, phenotypes not seen in mice lackingFgfr3. These data demonstrate that FGF18 signals through another FGFR to regulate osteoblast growth. Signaling to multiple FGFRs positions FGF18 to coordinate chondrogenesis in the growth plate with osteogenesis in cortical and trabecular bone.

Prognostics, such as remaining useful life (RUL) prediction, is a crucial task in condition-based maintenance. A major challenge in data-driven prognostics is the difficulty of obtaining a sufficient number of samples of failure progression. However, for traditional machine learning methods and deep neural networks, enough training data is a prerequisite to train good prediction models. In this work, we proposed a transfer learning algorithm based on Bi-directional Long Short-Term Memory (BLSTM) recurrent neural networks for RUL estimation, in which the models can be first trained on different but related datasets and then fine-tuned by the target dataset. Extensive experimental results show that transfer learning can in general improve the prediction models on the dataset with a small number of samples. There is one exception that when transferring from multi-type operating conditions to single operating conditions, transfer learning led to a worse result.

Currently gear fault diagnosis is mainly based on vibration signals with a few studies on acoustic signal analysis. However, vibration signal acquisition is limited by its contact measuring while traditional acoustic-based gear fault diagnosis relies heavily on prior knowledge of signal processing techniques and diagnostic expertise. In this paper, a novel deep learning-based gear fault diagnosis method is proposed based on sound signal analysis. By establishing an end-to-end convolutional neural network (CNN), the time and frequency domain signals can be fed into the model as raw signals without feature engineering. Moreover, multi-channel information from different microphones can also be fused by CNN channels without using an extra fusion algorithm. Our experiment results show that our method achieved much better performance on gear fault diagnosis compared with other traditional gear fault diagnosis methods involving feature engineering. A publicly available sound signal dataset for gear fault diagnosis is also released and can be downloaded as instructed in the conclusion section.

Interactions between human leukocyte antigens (HLAs) and peptides play a critical role in the human immune system. Accurate computational prediction of HLA-binding peptides can be used for peptide drug discovery. Currently, the best prediction algorithms are neural network-based pan-specific models, which take advantage of the large amount of data across HLA alleles. However, current pan-specific models are all based on the pseudo sequence encoding for modeling the binding context, which is based on 34 positions identified from the HLA protein-peptide bound structures in early works. In this work, we proposed a novel deep convolutional neural network model (DCNN) for HLA-peptide binding prediction, in which the encoding of the HLA sequence and the binding context are both learned by the network itself without requiring the HLA-peptide bound structure information. Our DCNN model is also characterized by its binding context extraction layer and dual outputs with both binding affinity output and binding probability outputs. Evaluation on public benchmark datasets shows that our DeepSeqPan model without HLA structural information in training achieves state-of-the-art performance on a large number of HLA alleles with good generalization capability. Since our model only needs raw sequences from the HLA-peptide binding pairs, it can be applied to binding predictions of HLAs without structure information and can also be applied to other protein binding problems such as protein-DNA and protein-RNA bindings. The implementation code and trained models are freely available at https://github.com/pcpLiu/DeepSeqPan.

Human leukocyte antigen (HLA) complex molecules play an essential role in immune interactions by presenting peptides on the cell surface to T cells. With significant deep learning progress, a series of neural network-based models have been proposed and demonstrated with their excellent performances for peptide-HLA class I binding prediction. However, there is still a lack of effective binding prediction models for HLA class II protein binding with peptides due to its inherent challenges. We present a novel sequence-based pan-specific neural network structure, DeepSeaPanII, for peptide-HLA class II binding prediction in this work. Our model is an end-to-end neural network model without the need for pre-or post-processing on input samples compared with existing pan-specific models. Besides state-of-the-art performance in binding affinity prediction, DeepSeqPanII can also extract biological insight on the binding mechanism over the peptide by its attention mechanism-based binding core prediction capability. The leave-one-allele-out cross-validation and benchmark evaluation results show that our proposed network model achieved state-of-the-art performance in HLA-II peptide binding. The source code and trained models are freely available at https://github.com/pcpLiu/DeepSeqPanII .

Convolutional neural networks (CNN) have been shown to outperform conventional methods in DNA-protien binding specificity prediction. However, whether we can transfer this success to protien-peptide binding affinity prediction depends on appropriate design of the CNN architectue that calls for thorough understanding how to match the architecture to the problem. Here we propose DeepMHC, a deep convolutional neural network (CNN) based protein-peptide binding prediction algorithm for achieving better performance in MHC-I peptide binding affinity prediction than conventional algorithms. Our model takes only raw binding peptide sequences as input without needing any human-designed features and othe physichochemical or evolutionary information of the amino acids. Our CNN models are shown to be able to learn non-linear relationships among the amino acid positions of the peptides to achieve highly competitive performance on most of the IEDB benchmark datasets with a single model architecture and without using any consensus or composite ensemble classifier models. By systematically exploring the best CNN architecture, we identified critical design considerations in CNN architecture development for peptide-MHC binding prediction.