Dictionary learning (DL), implemented via matrix factorization (MF), is commonly used in computational biology to tackle ubiquitous clustering problems. The method is favored due to its conceptual simplicity and relatively low computational complexity. However, DL algorithms produce results that lack interpretability in terms of real biological data. Additionally, they are not optimized for graph-structured data and hence often fail to handle them in a scalable manner. In order to address these limitations, we propose a novel DL algorithm called online convex network dictionary learning (online cvxNDL). Unlike classical DL algorithms, online cvxNDL is implemented via MF and designed to handle extremely large datasets by virtue of its online nature. Importantly, it enables the interpretation of dictionary elements, which serve as cluster representatives, through convex combinations of real measurements. Moreover, the algorithm can be applied to data with a network structure by incorporating specialized subnetwork sampling techniques. To demonstrate the utility of our approach, we apply cvxNDL on 3D-genome RNAPII ChIA-Drop data with the goal of identifying important long-range interaction patterns (long-range dictionary elements). ChIA-Drop probes higher-order interactions, and produces data in the form of hypergraphs whose nodes represent genomic fragments. The hyperedges represent observed physical contacts. Our hypergraph model analysis has the objective of creating an interpretable dictionary of long-range interaction patterns that accurately represent global chromatin physical contact maps. Through the use of dictionary information, one can also associate the contact maps with RNA transcripts and infer cellular functions. To accomplish the task at hand, we focus on RNAPII-enriched ChIA-Drop data from Drosophila Melanogaster S2 cell lines. Our results offer two key insights. First, we demonstrate that online cvxNDL retains the accuracy of classical DL (MF) methods while simultaneously ensuring unique interpretability and scalability. Second, we identify distinct collections of proximal and distal interaction patterns involving chromatin elements shared by related processes across different chromosomes, as well as patterns unique to specific chromosomes. To associate the dictionary elements with biological properties of the corresponding chromatin regions, we employ Gene Ontology (GO) enrichment analysis and perform multiple RNA coexpression studies.

This paper studies the global energy efficiency (GEE) optimization problem in a typical heterogeneous network (HetNet), where a macro base station (BS) and multiple micro BSs share the same spectrum band. Conventional optimization methods typically involve collecting global instantaneous channel state information (CSI) and utilizing centralized optimization algorithms to obtain the optimal transmit power and enhance the GEE. However, it is expensive to obtain the global instantaneous CSI in practical scenarios, and the optimization algorithms tend to be time-consuming. In this paper, we develop a lightweight cloud-edge collaboration framework based on the deep reinforcement learning (DRL) technique, such that the BSs in the edge do not need to exchange local instantaneous information with each other, and the core network in the cloud can collect only historical data rates and energy consumption information from each edge BS and feeds the calculated global reward back to them. Within the frame-work, we establish an independent actor-critic structure for each BS, and design a multi-agent independent actor-critic (MAIAC) power control algorithm, which enables each BS to determine its transmit power locally and enhance the GEE based on only local information. Simulation results indicate that the proposed MAIAC algorithm can achieve comparable GEE performance with conventional algorithms while requiring significantly less time complexity.

We consider the problem of determining the mutational support and distribution of the SARS-CoV-2 viral genome in the small-sample regime. The mutational support refers to the unknown number of sites that may eventually mutate in the SARS-CoV-2 genome while mutational distribution refers to the distribution of point mutations in the viral genome across a population. The mutational support may be used to assess the virulence of the virus and guide primer selection for real-time RT-PCR testing. Estimating the distribution of mutations in the genome of different subpopulations while accounting for the unseen may also aid in discovering new variants. To estimate the mutational support in the small-sample regime, we use GISAID sequencing data and our state-of-the-art polynomial estimation techniques based on new weighted and regularized Chebyshev approximation methods. For distribution estimation, we adapt the well-known Good-Turing estimator. Our analysis reveals several findings: First, the mutational supports exhibit significant differences in the ORF6 and ORF7a regions (older versus younger patients), ORF1b and ORF10 regions (females versus males) and in almost all ORFs (Asia/Europe/North America). Second, even though the N region of SARS-CoV-2 has a predicted 10% mutational support, mutations fall outside of the primer regions recommended by the CDC.

We consider a typical heterogeneous network (HetNet), which consists of a macro base station (BS) and multiple small BSs sharing the same spectrum band. Since the spectrum sharing among different BS-user links may cause severe mutual interference and degrades the global energy efficiency (GEE), it is important to optimize the transmit power of each BS and enhance the GEE. Conventional methods first collect the global instantaneous channel state information (CSI) and then optimize the transmit power in a centralized manner. Nevertheless, it is demanding to obtain the global instantaneous CSI in practical situations and the centralized optimization may easily overwhelm the coherence time of wireless channels. To tackle these issues, we leverage the strong computing capability of the (cloud) core network and the fast configuration capability of (edge) BSs and propose an intelligent cloud-edge collaboration framework. By properly designing the cloud-edge collaboration, we develop a deep reinforcement learning (DRL) based energy efficient power control algorithm. With the proposed algorithm, each BS can configure its transmit power independently and enhance the GEE. Simulation results reveal that, in both static-user and mobile-user scenarios, the proposed algorithm can provide comparable GEE performance with the conventional methods while requiring a much lower time complexity.

In a typical heterogeneous network (HetNet), in which a macro base station (BS) and multiple small BSs coexist on the same spectrum band, energy-efficiency (EE) performance is an important design metric and is highly related to the transmit power of BSs. Conventional methods optimize BSs’ transmit power to enhance the EE by assuming that the global channel state information (CSI) is available. However, it is challenging or expensive to collect the instantaneous global CSI in the HetNet. In this paper, we utilize deep reinforcement learning (DRL) technique to design an intelligent power control algorithm, with which each BS can independently determine the transmit power based on only local information. Simulation results demonstrate that the proposed algorithm outperforms conventional methods in terms of both EE performance and time complexity.

Generative graph models create instances of graphs that mimic the properties of real-world networks. Generative models are successful at retaining pairwise associations in the underlying networks but often fail to capture higher-order connectivity patterns known as network motifs. Different types of graphs contain different network motifs, an example of which are triangles that often arise in social and biological networks. It is hence vital to capture these higher-order structures to simulate real-world networks accurately. We propose Multi-MotifGAN (MMGAN), a motif-targeted Generative Adversarial Network (GAN) that generalizes the benchmark NetGAN approach. The generalization consists of combining multiple biased random walks, each of which captures a different motif structure. MMGAN outperforms NetGAN at creating new graphs that accurately reflect the network motif statistics of input graphs such as Citeseer, Cora and Facebook.

Matrix factorization (MF) is a versatile learning method that has found wide applications in various data-driven disciplines. Still, many MF algorithms do not adequately scale with the size of available datasets and/or lack interpretability. To improve the computational efficiency of the method, an online (streaming) MF algorithm was proposed in Mairal et al., 2010. To enable data interpretability, a constrained version of MF, termed convex MF, was introduced in Ding et al., 2010. In the latter work, the basis vectors are required to lie in the convex hull of the data samples, thereby ensuring that every basis can be interpreted as a weighted combination of data samples. No current algorithmic solutions for online convex MF are known as it is challenging to find adequate convex bases without having access to the complete dataset. We address both problems by proposing the first online convex MF algorithm that maintains a collection of constant-size sets of representative data samples needed for interpreting each of the basis (Ding et al., 2010) and has the same almost sure convergence guarantees as the online learning algorithm of Mairal et al., 2010. Our proof techniques combine random coordinate descent algorithms with specialized quasi-martingale convergence analysis. Experiments on synthetic and real world datasets show significant computational savings of the proposed online convex MF method compared to classical convex MF. Since the proposed method maintains small representative sets of data samples needed for convex interpretations, it is related to a body of work in theoretical computer science, pertaining to generating point sets (Blum et al., 2016), and in computer vision, pertaining to archetypal analysis (Mei et al., 2018). Nevertheless, it differs from these lines of work both in terms of the objective and algorithmic implementations.

Motivation: DNA methylation is one of the most important epigenetic mechanisms in cells that exhibits a significant role in controlling gene expressions. Abnormal methylation patterns have been associated with cancer, imprinting disorders and repeat-instability diseases. As inexpensive bisulfite sequencing approaches have led to significant efforts in acquiring methylation data, problems of data storage and management have become increasingly important. The de facto compression method for methylation data is gzip, which is a general purpose compression algorithm that does not cater to the special format of methylation files. We propose METHCOMP, a new compression scheme tailor-made for bedMethyl files, which supports random access. Results: We tested the METHCOMP algorithm on 24 bedMethyl files retrieved from four randomly selected ENCODE assays. Our findings reveal that METHCOMP offers an average compression ratio improvement over gzip of up to 7.5x. As an example, METHCOMP compresses a 48 GB file to only 0.9 GB, which corresponds to a 98% reduction in size. Availability and implementation: METHCOMP is freely available at https://github.com/jianhao2016/METHCOMP. Supplementary information: Supplementary data are available at Bioinformatics online.