DeepPlnc: Discovering plant lncRNAs through multimodal deep learning on sequential data

Various noncoding elements of genome have gained attention for their regulatory roles where the lncRNAs are very recent and most intriguing for their possible functions. Due to limited information about lncRNAs, their characterization remains a big challenge, especially in plants. Plant lncRNAs differ a lot from others even in the mode of transcription and display poor sequence conservation. Scarce resources exist to annotate for lncRNAs with satisfactory reliability. Here, we present a deep learning approach-based software, DeepPlnc, to accurately identify plant lncRNAs across the plant genomes. DeepPlnc, unlike most of the existing software, can even accurately annotate the incomplete length transcripts also which are very common in de novo assembled transcriptomes. It has incorporated a bi-modal architecture of Convolution Neural Nets while extracting information from the sequences of nucleotides and secondary structure representations for plant lncRNAs. DeepPlnc scored high on all the considered performance metrics while breaching the average accuracy of >95% when tested across different experimentally validated datasets. The software was comprehensively benchmarked against some of the recently published tools to identify the plant lncRNAs where it consistently outperformed all the compared tools for all the performance metrics and for all the considered benchmarking datasets. DeepPlnc will be an important resource for reference free identification and annotation of transcriptome and genome for lncRNAs in plants. DeepPlnc has been made freely available as a web-server at https://scbb.ihbt.res.in/DeepPlnc/. Besides this, a stand-alone version is also provided at GitHub at https://github.com/SCBB-LAB/DeepPlnc/.

Download Full-text

De Novo Prediction of RNA 3D Structures with Deep Learning

10.1101/2021.08.30.458226 ◽

2021 ◽

Author(s):

Julius Ramakers ◽

Christopher Frederik Blum ◽

Sabrina König ◽

Stefan Harmeling ◽

Markus Kollmann

Keyword(s):

Monte Carlo ◽

Deep Learning ◽

Nucleic Acid ◽

Predictive Power ◽

De Novo ◽

Learning Approach ◽

Tree Search ◽

Blind Test ◽

Monte Carlo Tree Search ◽

Score Model

We present a Deep Learning approach to predict 3D folding structures of RNAs from their nucleic acid sequence. Our approach combines an autoregressive Deep Generative Model, Monte Carlo Tree Search, and a Score Model to find and rank the most likely folding structures for a given RNA sequence. We confirm the predictive power of our approach by setting new benchmarks for some longer sequences in a simulated blind test of the RNA Puzzles prediction challenge.

Download Full-text

Classification of Hemodynamics Scenarios from a Public Radar Dataset Using a Deep Learning Approach

Sensors ◽

10.3390/s21051836 ◽

2021 ◽

Vol 21 (5) ◽

pp. 1836

Author(s):

Gašper Slapničar ◽

Wenjin Wang ◽

Mitja Luštrek

Keyword(s):

Deep Learning ◽

Valsalva Maneuver ◽

Learning Approach ◽

Network Architectures ◽

Breathing Rate ◽

Limited Information ◽

Data Types ◽

Input Signals ◽

Radar Sensing ◽

Contact Free

Contact-free sensors offer important advantages compared to traditional wearables. Radio-frequency sensors (e.g., radars) offer the means to monitor cardiorespiratory activity of people without compromising their privacy, however, only limited information can be obtained via movement, traditionally related to heart or breathing rate. We investigated whether five complex hemodynamics scenarios (resting, apnea simulation, Valsalva maneuver, tilt up and tilt down on a tilt table) can be classified directly from publicly available contact and radar input signals in an end-to-end deep learning approach. A series of robust k-fold cross-validation evaluation experiments were conducted in which neural network architectures and hyperparameters were optimized, and different data input modalities (contact, radar and fusion) and data types (time and frequency domain) were investigated. We achieved reasonably high accuracies of 88% for contact, 83% for radar and 88% for fusion of modalities. These results are valuable in showing large potential of radar sensing even for more complex scenarios going beyond just heart and breathing rate. Such contact-free sensing can be valuable for fast privacy-preserving hospital screenings and for cases where traditional werables are impossible to use.

Download Full-text

Comparison of various Activation Functions A Deep Learning Approach

International Journal of Computer Sciences and Engineering ◽

10.26438/ijcse/v6i3.122126 ◽

2018 ◽

Vol 6 (3) ◽

pp. 122-126

Author(s):

Mohammed Ibrahim Khan ◽

◽

Akansha Singh ◽

Anand Handa ◽

◽

...

Keyword(s):

Deep Learning ◽

Learning Approach ◽

Activation Functions

Download Full-text

A Deep Learning based Arabic Script Recognition System: Benchmark on KHAT

The International Arab Journal of Information Technology ◽

10.34028/iajit/17/3/3 ◽

2020 ◽

Vol 17 (3) ◽

pp. 299-305 ◽

Cited By ~ 1

Author(s):

Riaz Ahmad ◽

Saeeda Naz ◽

Muhammad Afzal ◽

Sheikh Rashid ◽

Marcus Liwicki ◽

...

Keyword(s):

Deep Learning ◽

Character Recognition ◽

Data Augmentation ◽

Short Term Memory ◽

Recognition System ◽

Learning Approach ◽

Arabic Text ◽

Data Set ◽

Processing Step ◽

Handwritten Arabic

This paper presents a deep learning benchmark on a complex dataset known as KFUPM Handwritten Arabic TexT (KHATT). The KHATT data-set consists of complex patterns of handwritten Arabic text-lines. This paper contributes mainly in three aspects i.e., (1) pre-processing, (2) deep learning based approach, and (3) data-augmentation. The pre-processing step includes pruning of white extra spaces plus de-skewing the skewed text-lines. We deploy a deep learning approach based on Multi-Dimensional Long Short-Term Memory (MDLSTM) networks and Connectionist Temporal Classification (CTC). The MDLSTM has the advantage of scanning the Arabic text-lines in all directions (horizontal and vertical) to cover dots, diacritics, strokes and fine inflammation. The data-augmentation with a deep learning approach proves to achieve better and promising improvement in results by gaining 80.02% Character Recognition (CR) over 75.08% as baseline.

Download Full-text