iACP-MultiCNN: Multi-channel CNN based anticancer peptides identification.

Cancer is one of the most dangerous diseases in the world that often leads to misery and death. Current treatments include different kinds of anticancer therapy which exhibit different types of side effects. Because of certain physicochemical properties, anticancer peptides (ACPs) have opened a new path of treatments for this deadly disease. That is why a well-performed methodology for identifying novel anticancer peptides has great importance in the fight against cancer. In addition to the laboratory techniques, various machine learning and deep learning methodologies have developed in recent years for this task. Although these models have shown reasonable predictive ability, there's still room for improvement in terms of performance and exploring new types of algorithms. In this work, we have proposed a novel multi-channel convolutional neural network (CNN) for identifying anticancer peptides from protein sequences. We have collected data from the existing state-of-the-art methodologies and applied binary encoding for data preprocessing. We have also employed k-fold cross-validation to train our models on benchmark datasets and compared our models' performance on the independent datasets. The comparison has indicated our models' superiority on various evaluation metrics. We think our work can be a valuable asset in finding novel anticancer peptides. We have provided a user-friendly web server for academic purposes and it is publicly available at: http://103.99.176.239/iacp-cnn/.

Bioacoustic classification of avian calls from raw sound waveforms with an open-source deep learning architecture.

The use of autonomous recordings of animal sounds to detect species is a popular conservation tool, constantly improving in fidelity as audio hardware and software evolves. Current classification algorithms utilise sound features extracted from the recording rather than the sound itself, with varying degrees of success. Neural networks that learn directly from the raw sound waveforms have been implemented in human speech recognition but the requirements of detailed labelled data have limited their use in bioacoustics. Here we test SincNet, an efficient neural network architecture that learns from the raw waveform using sinc-based filters. Results using an off-the-shelf implementation of SincNet on a publicly available bird sound dataset (NIPS4Bplus) show that the neural network rapidly converged reaching accuracies of over 65% with limited data. Their performance is comparable with traditional methods after hyperparameter tuning but they are more efficient. Learning directly from the raw waveform allows the algorithm to select automatically those elements of the sound that are best suited for the task, bypassing the onerous task of selecting feature extraction techniques and reducing possible biases. We use publicly released code and datasets to encourage others to replicate our results and to apply SincNet to their own datasets; and we review possible enhancements in the hope that algorithms that learn from the raw waveform will become useful bioacoustic tools.

Urbanization-globalization-CO2 emissions nexus revisited: empirical evidence from South Africa.

The environmental effects of urbanization and globalization are still subject to debate among scholars. South Africa is the most globalized, most urbanized and the most carbon-intensive economy in Sub Saharan Africa (SSA) region. Taking this into cognizance, this study examines the effects of urbanization and globalization on CO2 emissions for South Africa using time series annual data for the period 1980-2017. Zivot and Andrews single and Bai and Perron multiple structural break unit root tests are employed to assess if all the series are stationary. This procedure follows ARDL cointegration test to check the presence of a long-run association among variables. Having been confirmed about such a cointegrating relation, ARDL short-run and long run coefficients indicate that urbanization induces CO2 emissions while only long-run significant emissions effect of globalization was noted. Toda-Yamamoto non-causality test reports a bi-directional causal link between urbanization and CO2 emissions. No causal link is observed between globalization and CO2 emissions. Variance decomposition results do not rule out these effects in future. Policy implications are discussed.