De novo protein sequencing is one of the key problems in mass spectrometry-based proteomics, especially for novel proteins such as monoclonal antibodies for which genome information is often limited or not available. However, due to limitations in peptides fragmentation and coverage, as well as ambiguities in spectra interpretation, complete de novo assembly of unknown protein sequences still remains challenging. To address this problem, we propose an integrated system, ALPS, which for the first time can automatically assemble full-length monoclonal antibody sequences. Our system integrates de novo sequencing peptides, their quality scores and error-correction information from databases into a weighted de Bruijn graph to assemble protein sequences. We evaluated ALPS performance on two antibody data sets, each including a heavy chain and a light chain. The results show that ALPS was able to assemble three complete monoclonal antibody sequences of length 216–441 AA, at 100% coverage, and 96.64–100% accuracy.

Liquid chromatography with tandem mass spectrometry (LC-MS/MS) based quantitative proteomics provides the relative different protein abundance in healthy and disease-afflicted patients, which offers the information for molecular interactions, signaling pathways, and biomarker identification to serve the drug discovery and clinical research. Typical analysis workflow begins with the peptide feature detection and intensity calculation from LC-MS map. We are the first to propose a deep learning based model, DeepIso, that combines recent advances in Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) to detect peptide features of different charge states, as well as, estimate their intensity. Existing tools are designed with limited engineered features and domain-specific parameters, which are hardly updated despite a huge amount of new coming proteomic data. On the other hand, DeepIso consisting of two separate deep learning based modules, learns multiple levels of representation of high dimensional data itself through many layers of neurons, and adaptable to newly acquired data. The peptide feature list reported by our model matches with 97.43% of high quality MS/MS identifications in a benchmark dataset, which is higher than the matching produced by several widely used tools. Our results demonstrate that novel deep learning tools are desirable to advance the state-of-the-art in protein identification and quantification.

Handwritten character recognition complexity varies among different languages due to distinct shapes, strokes and number of characters. Numerous works in handwritten character recognition are available for English with respect to other major languages such as Bangla. Existing methods use distinct feature extraction techniques and various classification tools in their recognition schemes. Recently, Convolutional Neural Network (CNN) is found efficient for English handwritten character recognition. In this paper, a CNN based Bangla handwritten character recognition is investigated. The proposed method normalizes the written character images and then employ CNN to classify individual characters. It does not employ any feature extraction method like other related works. 20000 handwritten characters with different shapes and variations are used in this study. The proposed method is shown satisfactory recognition accuracy and outperformed some other prominent exiting methods.

Handwritten recognition has drawn profound attention since decades ago due to its numerous potential applications in real life. Research on unconstrained handwritten recognition in some languages has achieved attractive advancement, but it lags behind for Bengali even though it is the major language spoken by about 230 million people in the Indian subcontinent, and even the first and official language of Bangladesh. Recently, the use of convolutional neural network (CNN) has been reported with high accuracy in pattern recognition and computer vision problems. The main purpose of this study is to provide an architecture of a CNN to improve the accuracy of handwritten Bengali numerals recognition (HBNR) and compare its performance with the existing ones. We proposed a new CNN architecture, VGG-11M, which improves an existing one (VGG-11). The normalized and rescaled images of each numeral were augmented by different transformation operations to increase the training samples and to add diversity in the dataset. Then, the images were used to train the proposed VGG-11M model. The recognition accuracy of the developed system was tested on both training and test sets of three publicly available handwritten Bengali numerals database at different resolutions. Finally the performance of the model was compared with four other architectures (LeNet-5, ResNet-50, VGG-11, and VGG-16). The highest accuracy 99.80%, 99.66%, and 99.25% was obtained using the proposed architecture on the test set of ISI, CMATERDB, and NUMTADB dataset, respectively, at resolution 32 × 32. The proposed VGG-11M outperformed the existing architectures of CNN on HBNR.

No abstract

A promising technique of discovering disease biomarkers is to measure the relative protein abundance in multiple biofluid samples through liquid chromatography with tandem mass spectrometry (LC-MS/MS) based quantitative proteomics. The key step involves peptide feature detection in the LC-MS map, along with its charge and intensity. Existing heuristic algorithms suffer from inaccurate parameters and human errors. As a solution, we propose PointIso, the first point cloud based arbitrary-precision deep learning network to address this problem. It consists of attention based scanning step for segmenting the multi-isotopic pattern of 3D peptide features along with the charge, and a sequence classification step for grouping those isotopes into potential peptide features. PointIso achieves 98% detection of high-quality MS/MS identified peptide features in a benchmark dataset. Next, the model is adapted for handling the additional ‘ion mobility’ dimension and achieves 4% higher detection than existing algorithms on the human proteome dataset. Besides contributing to the proteomics study, our novel segmentation technique should serve the general object detection domain as well.

No abstract

Here we present GlycanFinder, a database search and de novo sequencing tool for the analysis of intact glycopeptides from mass spectrometry data. GlycanFinder integrates peptide-based and glycan-based search strategies to address the challenge of complex fragmentation of glycopeptides. A deep learning model is designed to capture glycan tree structures and their fragment ions for de novo sequencing of glycans that do not exist in the database. We performed extensive analyses to validate the false discovery rates (FDRs) at both peptide and glycan levels and to evaluate GlycanFinder based on comprehensive benchmarks from previous community-based studies. Our results show that GlycanFinder achieved comparable performance to other leading glycoproteomics softwares in terms of both FDR control and the number of identifications. Moreover, GlycanFinder was also able to identify glycopeptides not found in existing databases. Finally, we conducted a mass spectrometry experiment for antibody N-linked glycosylation profiling that could distinguish isomeric peptides and glycans in four immunoglobulin G subclasses, which had been a challenging problem to previous studies.