Dec. 2023, Volume 11 Issue 4

Download cover

Sep. 2023, Volume 11 Issue 3

The rapid development of biological technology (BT) and information technology (IT) especially of genomics and artificial intelligence (AI) is bringing great potential for revolutionizing future medicine. Zhang et al. proposed the concept and framework of Digital Life Systems or dLife as a new paradigm to unleash this potential. The framework aims to build cyber twins of healthy or diseased human body through quantitative measurement, informatic representation, and mathematical/computational modeling. The cyber twins can be used to simulate complex biological processes and deduce effects of medical treatments. dLife is the route toward future AI precision medicine and should be the new paradigm for future biological and medical research. For details please refer to the article by Zhang et al. in pp. 207—213. Download cover

Jun. 2023, Volume 11 Issue 2

Large language model-based chatbots present novel possibilities and challenges for bioinformatics. The vivid imagery depicted on the cover portrays students engaged an exciting conversation with a chatbot at sunrise, symbolizing the dawn of new prospects in bioinformatics. However, the mist enveloping the mountain valleys serves as a metaphor for the challenges that accompany these opportunities. Positioned alongside the students, an experienced educator devises strategies to embrace the possibilities while mitigating associated risks. One such approach is the OPTIMAL model proposed in pages 105-108, which serves as a guiding framework to facilitate communication with chatbots. By leveraging this model, students can enhance their coding skills, prompting skills, and critical thinking, paving the way for a transformative bioinformatics learning experience.

Mar. 2023, Volume 11 Issue 1

Somatic mutation is considered the primary direct cause of tumorigenesis. A cancer patient can carry up to tens of thousands of single base substitution somatic mutations. Like the Butterfly effect, these somatic mutations lead to a series of regulatory function changes, including gene expression dysregulation. Jiang et al. hypothesized that associations between gene expression and somatic mutations can be translated into the associations between gene expression and mutational signatures for showing the regulation effects between somatic mutation and gene expression dysregulation. And results based on 49 established mutational signatures and 33 cancer types demonstrate that the hypothesis is reasonable. This work reveals the key message that some somatic mutations can lead to gene expression alteration consequences. By modeling these associations between mutational signature and gene expression, mutational signatures can be potentially inferred in future RNA-seq studies. For details please refer to the article by Jiang et al. in pp. 31-43.

Dec. 2022, Volume 10 Issue 4

Pooled CRISPR screen is a promising tool in drug targets or essential genes identification combined with different phenotype readouts. Aside from continuous improvements in technology, more and more bioinformatics methods have been developed to analyze the data obtained by CRISPR screens which facilitate better understanding of physiological effects. In this issue, Zhao et al. provide an overview on the application of CRISPR screens and bioinformatics approaches to analyzing different types of CRISPR screen data. Mechanisms and underlying challenges for the analysis of dropout screens, sorting-based screens and single-cell screens are also discussed. For details please refer to the article by Zhao et al. in pp. 307-320. We would like to thank Biorender.com for providing models of scientific illustrations.

Sep. 2022, Volume 10 Issue 3

Superpixel segmentation, an image over-segment technique, can provide boundary-priors for medical image applications to reduce annotation costs and improve segmentation accuracy. Traditionally, size uniformity is one of the significant features of superpixels. However, in medical images, where objects’ scale varies greatly and background regions are often flat, size uniformity rarely conforms to the varying content. To free users from the predicament of setting a suitable superpixel size for a given application, we propose a scale-adaptive superpixel algorithm for medical images. It generates superpixels with multi-scale according to the richness of image content, based on a new path-based distance measure and region-growth model. For details please refer to the article by Sun et al. in pp. 264‒275. We acknowledge Limin Sun and Yuanfeng Zhou for the cover design.

Jun. 2022, Volume 10 Issue 2

Precise prediction of the trend of the pandemic has been crucial in COVID-19, as it allows the governments and healthcare systems to effectively distribute and prioritize resource allocations. In this issue, Jo et al. present a neural network model to make quantile predictions on COVID-19 death tolls at the county level. This fine geographical scale, which is rare in publicly available models, can benefit state-level officials to coordinate resources within the state. The quantile predictions also inform people about the distribution of the predicted death tolls, allowing better evaluation of possible trajectories of the severity. The model’s scalability and generalizability further suggests its application in generating other useful predictions such as new cases or hospitalizations by incorporating additional data sources. For details please refer to the article by Jo et al. in pp. 125‒138.

Mar. 2022, Volume 10 Issue 1

Alternative polyadenylation (APA) plays an essential role in post-transcriptional regulation for most human genes and mostly function within 3′UTR regions. This cover image shows the genetic variants associated with alternative mRNA 3′UTR lengths. A change from the T allele to the C allele leads to the lengthening of a 3′UTR with distinct disease status. These 3′UTR associated genetic variants provide substantial new insights into the molecular mechanisms underlying many human complex traits and diseases. For details please refer to the article by Li et al. in pp. 44‒54. We acknowledge Xiangli Liu for the cover design.

Dec. 2021, Volume 9 Issue 4

Like oil blending into larger liquid blobs, the contents of cells can also be separated into droplets. The transformation of a single-phase system into a multiple-phase system is defined as phase separation. Phase separation is one of biology’s hottest questions in recent years. Numerous physical theories and biological experiments have been developed to uncover the underlying principles of phase separation in biology. Whether a solution is able to undergo phase separation in a cell, and whether phase separation is nucleation-limited or diffusion-limited depend strongly on the component concentration and temperature. Based on these physical theories, the occurrences of some diseases could be understood and synthetic organelles could be designed to implement some specific functions. For details please refer to the article by Shi et al. in pp. 378‒399. Download cover

Sep. 2021, Volume 9 Issue 3

The year 2021 is the 20th anniversary of the publication of the draft human genome. The sequencing of the human genome has brought life sciences into a new era, the era to understand the information systems of life in a quantitative manner. Quantitative Biology publishes a series of featured articles to celebrate the anniversary. Download cover

Jun. 2021, Volume 9 Issue 2

GWAS have identified many genetic variants associated with increased risk of Alzheimer’s disease (AD). These susceptibility loci may affect AD indirectly through a combination of physiological brain changes, which are detectable via magnetic resonance imaging. In this issue, Knutson and Pan examined the effects of brain imaging derived phenotypes with genetic etiology on AD, comparing the following summary statistic based methods: two-sample Mendelian randomization, generalized summary statistics based Mendelian randomization, transcriptome wide association studies, and the adaptive sum of powered score test. Using publicly available GWAS datasets from the IGAP and UK Biobank, they identified 35 IDPs possibly associated with AD, many of which have well established or biologically plausible links to the characteristic cognitive impairments of AD. These results highlight the increased power for detecting genetic associations achieved by multiple correlated SNP-based methods over MR methods based on independent SNPs. For details please refer to the article by Knutson and Pan in pp. 185‒200. We acknowledge Marni Kaldjian for making the cover image. Download cover

Mar. 2021, Volume 9 Issue 1

Coronavirus SARS-CoV-2 infects host cells via binding to the cell receptor ACE2 with its spike proteins. Atomistic molecular dynamics simulations reveal that the receptor binding domain (RBD) of spike protein can tightly bind to both open and closed ACE2 receptors, mostly by specific interactions between RBD and ACE2 N-terminal helices. The water molecules residing at the RBD-ACE2 interface play critical roles in stabilizing the complex structure: on average about 15 water molecules are observed at the RBD-ACE2 interface. Some water molecules stay at the interface over 10 nanoseconds, suggesting that the significant contribution to the RBD binding. Engineered ACE2 proteins or peptides can be potential pharmaceutical molecules to interfere the infection of SARS-CoV-2. For details please refer to the article by Lupala et al. in pp. 61‒72. We acknowledge Letpub for its professional scientific illustration service. Download cover Download table of contents

Dec. 2020, Volume 8 Issue 4

Histone modifications play an important role in defining chromatin states and regulating gene expression. Many histone modifications, such as H3K27me3 and H3K9me3, form broad domains in the genome. Measured by ChIP-seq, board histone domains look like mountain ridges in the data and are more difficult to identify than sharp peaks. In this issue, Zang et al. present a computational method, RECOGNICER, for identifying cross-scale board domains from ChIP-seq data using a coarse-graining approach. For details please refer to the article by Zang et al. in pp. 359‒368. We acknowledge Yixuan Ren for making the cover image.

Sep. 2020, Volume 8 Issue 3

Microscopic imaging of live cells is routine in biology labs. Quantitative measurement requires precise recognition, labeling and tracking of each individual cell. This segmentation process is often quite challenging since images vary greatly in their features and qualities. Current segmentation methods rely on single boundary features, and are thus hardly applicable to different situations and shared between labs. In this issue, Ren et al. develop a robust and accurate cell segmentation program – Cellbow, using a deep neural network combined with a multi-layer training set strategy. It broadly segments fluorescent images of diverse cell types and overcomes the inhomogeneous foci in bright-field images. The software is available online. For details please refer to the article by Ren et al. in pp. 245‒255.
We acknowledge Nan Sheng for making the cover image.

Jun. 2020, Volume 8 Issue 2

RNA binding proteins (RBPs) are known as key post-transcriptional regulators. The recent technology, cross-linking and immunoprecipitation followed by sequencing (CLIP-seq), has made it possible to investigate the interaction between RBPs and RNAs. However, the association between the function and the binding of RBPs has not been systematically studied. In this issue, Lin and Ouyang present a large-scale analysis on the functional targets of human RBPs based on the enhanced CLIP-seq datasets. Their study uncovers that the translation termination site and the 3′ untranslated region are important binding positions of RNA decay-related RBPs and the regulation follows a cell-type specific pattern. It provides novel insights on the post-transcriptional mechanisms of RBPs. For details please refer to the article by Lin and Ouyang in pp. 119‒129. We acknowledge Matt Wimsatt in JAX Creative for making the cover image.

Mar. 2020, Volume 8 Issue 1

The CRISPR/Cas9 system has shown great potential in functional genomic screening by introducing short indels in protein-coding genes. However, short indel is usally not sufficient to generate loss-of-function of non-coding genomic element. In this issue, Tao et al. propose a strategy to construct a library of paired sgRNA expressing plasmids that can be used to efficiently generate chromosomal deletions, providing a scalable method for functional study of non-coding elements in mammalian cells. For details please refer to the article by Tao et al. in pp. 31-42.

Dec. 2019, Volume 7 Issue 4

Type III secretion system (T3SS) is a specialized protein delivery system in gram-negative bacteria, and the type III secreted effectors (T3SEs) play an important part in disease development for many plant and animal pathogens. Computational identification of T3SEs is a very challenging task in bioinformatics due to the lack of defined secretion signal and great sequence diversity. To exploit T3SE sequence information, Fu et al. employed a word embedding method to capture semantic correlations between amino acid fragments, which are complementary to position-specific features of sequence patterns commonly used in previous studies. Based on the combined feature representation and convolutional neural networks, they developed a new method, WEDeepT3, which achieved the state-of-the-art performance for the prediction of T3SEs. For details please refer to the article by Fu et al. in pp. 293-301.

Sep. 2019, Volume 7 Issue 3

Transcription factors (TF) regulate the expression level of targeted genes and furtherly effect biological functions. Song et al. developed EpiFIT, an online tool to infer functions of TF using sequence and epigenetic data. Through a series of examination experiments, they verified that EpiFIT can precisely interpret TF functions and build distal TF binding sites ? regulated genes associations with the help of epigenetic information. In a word, EpiFIT is a powerful tool for annotating the TF functions. They believe EpiFIT will facilitate the functional interpretation of other regulatory elements, and thus open a new door to understanding the regulatory mechanism. For details please refer to the article by Song et al. in pp. 233-243.

Jun. 2019, Volume 7 Issue 2

Competition among different types of biological molecules for limited resources is ubiquitous in biological processes. Molecular competition causes intricate behaviors in resource allocation, and thus introduces a hidden layer of regulatory mechanism by connecting components without direct physical interactions. Wei et al. built a unified coarse-grained competition motif model to quantitatively understand and predict diverse phenomena mediated by molecular competition. They systematically analyzed the properties of competing regulation from steady-state behavior to dynamic responses, evaluating how competition introduces indirect regulations and constraints among the targets, and how the existence of competitors could influence target-regulator response. The work provided a unified insight and a theoretical framework to understand and employ competition in both natural and synthetic systems. For details please refer to the article by Wei et al. in pp. 110-121.

Mar. 2019, Volume 7 Issue 1

Arachidonic acid network is a complex system with many pathways to which non-steroidal anti-inflammatroy drugs (NSAIDs) target. However, side effects have always been the disadvantages of these medicines. Using a natural 5-LOX inhibitor HOEC as probe, Yang et al. established a computational model to simulate the flux regulation of arachidonic acid network after HOEC treatment. There is balance in metabolic flux of every pathway, and with the reduce of flux in 5-LOX, the flux in COX pathway increases. The effect of HOEC on AA metabolic network mainly includes inhibiting the metabolic flux in 5-LOX pathway, and upregulating metabolic flux in COX pathway. The results of this study demonstrate that the dose-effect relationship of inhibitors targeting complex networks needs to be taken into account in the effect of the inhibitor on its target and impact on other targets in the network. For details please refer to the article by Yang et al. pp. 30-41.

Dec. 2018, Volume 6 Issue 4

The detection of protein complexes is a fundamental problem in proteomics and bioinformatics, which is equivalent to finding interesting sub-networks from protein-protein interaction networks. One such an interesting measure is the statistical significance of protein complexes in terms of P-values. However, how to evaluate the statistical significance of each detected protein complex has not received much attention in the literature. As a result, the statistical assessment of protein complexes still remains unsolved. To fulfill this void, a new yet simple P-value calculation method is presented, which is able to outperform existing methods on several benchmark protein-protein interaction networks. For details please refer to the article by Su et al. in pp. 313-320.

Sep. 2018, Volume 6 Issue 3

Alternative cleavage and polyadenylation (APA) results in mRNA isoforms with different 3′ UTR lengths. Significance analysis of alternative polyadenylation using RNA-seq (SAAP-RS) is a newly developed computational method that interrogates RNA-seq reads mapped to different 3′ UTR sequences (center) and analyze differential expression of 3′ UTR isoforms. SAAP-RS reveals 3′ UTR length differences among mouse brain cells, including neurons, astrocytes, oligodendrocytes, endothelial cells, and microglia. Among the cells, neurons display the longest 3′ UTRs, whereas microglia show the shortest ones. This finding is supported by single-cell sequencing data. SAAP-RS also reveals substantial 3′ UTR lengthening in human and mouse neurogenesis and its similarity to myogenesis. For details please refer to the article by Guvenek and Tian in pp. 253?266.

Jun. 2018, Volume 6 Issue 2

Networks are becoming more and more important common tools for representing biological systems. Decomposing such complex networks into module (or community) structure is a good way to find their underlying patterns. However, there are still few user-friendly tools to solve module detection in bipartite biological networks. BMTK is an online tool to effectively detect modules in such networks. It implements seven popular methods in a uniform platform and provides much convenience about graph visualization and results comparison among different methods. BMTK is a powerful tool for bipartite network module detection. Here BMTK is applied to a drug-target network to reveal its underlying distinct modular organization. For details please refer to the article by Wang et al. in pp. 186?192.

Mar. 2018, Volume 6 Issue 1

Sequence-specific transcription factors establish a diverse network of protein:DNA contacts to recognize target sites in the genome and fine-tune regulatory output. Closely related proteins within sub-families (e.g., OmpR-family paralogs, profiled in this issue) possess substantial structural and functional similarities, but key “specificity-determining residues” (SDRs) can introduce novel base preferences and global binding modes. This image displays the newly derived binding motifs of four E. coli winged helix-turn-helix transcription factors, surrounded by a subset of individual protein:DNA contacts extracted from X-ray crystal structures. Identifying potent SDRs from within this field of putative contacts is a major step toward the prediction of transcription factor binding sites, the design of synthetic regulatory proteins, and an understanding of regulatory evolution. For details please refer to the article by Joyce and Havranek in pp. 68-84.

Dec. 2017, Volume 5 Issue 4

Precision medicine attempts to tailor the right therapy for the right patient. However, it is still lack of powerful computational methods for an optimum target-drug recommendation. A novelty computation method for Precision Medicine Target-Drug Selection (PMTDS) is developed by Vasudevaraja et al. in this issue. It can priority the pair target-drug for individual style treatment of cancer based on genetic interaction networks and multi-omics data. Large-scale validation on adenocarcinoma (PDAC) tumors of the Cancer Genome Atlas (TCGA) shows that drugs selected by PMTDS have more sample-specific efficacy than the current clinical PDAC therapies. The PMTDS system provides an accurate and reliable source for target and off-label drug selection for precision cancer medicine. For details please refer to the article by Vasudevaraja et al. in pp. 380-394.

Aug. 2017, Volume 5 Issue 3

Biological studies are like fishing in the sea of information. We draw samples from the boundless sea using technologies like Hi-C, Chip-seq, RNA-seq and so on. Even with the high-throughput technologies, we could only see a small piece of the whole system. Computational methods are needed to filter the signal from noise and to obtain a complete picture of the system.

Jun. 2017, Volume 5 Issue 2

Monkey King with Golden Hoop: in Journey to the West, a classic Chinese mythological novel, a monkey called Sun Wukong helped his master Monk Tang Sanzang overcome various trials and tribulations during the pilgrimage for Buddhist scriptures. Although Sun Wukong possessed great power and talent, he cannot reach the final destination and get the real Buddhist scripture without Tang Sanzang’s control via the incantation of the golden hoop. The golden hoop is emphasized in this particular Monkey King to show its important role in this amazing adventure. Similarly, as Zhang et al. showed in this issue, the restricted Boltzmann machines (RBMs), though equipped with universal modeling power, cannot achieve satisfying modeling performance in the “p》N” scenario, e.g., cancer data analysis. To tackle this problem, Zhang et al. proposed a novel RBM called the elastic restricted Boltzmann machine (eRBM), in which additional elastic penalty term was introduced to control the model complexity while maintaining the modeling capacity. Extensive tests on both simulated data and real cancer profiling data were conducted to demonstrate the superiority of eRBMs over conventional methods when “p》N”. For details please refer to the article by Zhang et al. in pp. 159―172

Mar. 2017, Volume 5 Issue 1

The Yin and Yang of bacteria: the Chinese philosophical concept of Yin and Yang symbolizes how two opposing forces can harmoniously coincide. This particular Yin and Yang is made up of two types of bacteria, black and white, representing how competitive strains of bacteria can behave in paradoxically mutualistic ways. As Sadeghpour et al. show in this issue, if two strains of bacteria down-regulate transcription in one another, the resulting system can oscillate. As one strain wins transcriptionally, the other gains a growth advantage that allows it to eventually dominate the system. This result is in contrast to those obtained from studying co-repressive genes that often give rise to bistable gene expression patterns. Here, the growth and competition between the strains of the competitive consortium add a hidden layer of feedback that acts to create oscillations. For details please refer to the article by Sadeghpour et al. in pp. 55?66.

Dec. 2016, Volume 4 Issue 4

Single-cell RNA sequencing (scRNA-seq) is an emerging technology that enables high resolution detection of heterogeneities between cells. One important application of scRNA-seq data is to detect differential expression (DE) of genes between different types or subtypes of cells or between cells of different conditions. Currently, many researchers still use DE analysis methods developed for bulk RNA-Seq data on single-cell data. Some new methods for scRNA-seq data have also been developed. Bulk and single-cell RNA-seq data have quite different characteristics. In this issue, Miao et al. conducted a series of experiments on scRNA-seq data to systematically evaluate 14 popular DE analysis methods, including both traditional methods developed for bulk RNA-seq data and new methods specifically designed for scRNA-seq data. The comparison provides hints for choices of methods in different situations, and highlights the need for new methods based on better modeling of single-cell RNA-seq data. For details please refer to the article by Miao et al. in pages 243–260.

Sep. 2016, Volume 4 Issue 3

Chromatin interaction analysis by paired-end tag sequencing (ChIA-PET) technology was designed for detecting genome-wide chromatin loops mediated by a specific protein of interest, which has become one of the most important biological methods for understanding 3D genome organization. In this issue He et al. review five bioinformatics tools which are related to ChIA-PET data analysis and data mining. Meanwhile, they also introduce one interesting computational method which is to predict chromatin loops with ChIP-Seq data. The intention is to help reader to better understand ChIA-PET experiments and to select the most appropriate bioinformatics tools for their 3D genome research. For details please refer to the article by He et al. in pages 217―225.

May. 2016, Volume 4 Issue 2

The spatial organization of chromatins plays essential role in regulating transcriptional activity. There are mainly three kinds of methods to study 3D genome organization: molecular mapping, imaging and computational modeling. In this review we focus on reviewing (i) the recent developed super-resolution microscopy techniques to image and study chromatin, and (ii) the computational methods to analyze and visualize chromatin interactions derived from chromosome configuration capture (3C)-based molecular mapping methods, especially, from Hi-C data. The strategies developed from these different methods are not mutually exclusive —actually they are complemental to each other and can be combined together to study 3D genome organization.For details please refer to the article by Gao et al. in pages 129-147. The image was partially adapted from Ref. [11] with permission.

Mar. 2016, Volume 4 Issue 1

High-throughput technologies and computational methods are transforming biology from a qualitative, descriptive discipline into a quantitative, multi-parameter field. The wealth of publically available “big data” has promoted a paradigm shift in medical research. With the increase of integrative efforts across disciplines, a higher emphasis has been put on using data to define and drive hypotheses. Through a synergistic coupling of molecular indexes from transcriptomics, phenotypic traits from imaging, and clinical data from medical records, “radiotranscriptomics” can provide a keener insight into the molecular and functional alterations behind chronic and multifactorial disorders. We therefore propose “radiotranscriptomics” as a new member of the P4 medicine initiative, combining transcriptome information, including gene expression and isoform variation, and quantitative image annotations. For details please refer to the perspective by Amal Katrib, et al. in pages 1-12.

Jan. 2016, Volume 3 Issue 4

Computational tools for themodeling of protein structures as well as for structure-based protein design rely on models to quantify the respective molecular interactions from the atomic to the amino-acid residue levels. For this purpose, both physics-based models and statistical models, namely, models derived from statistics of experimentally determined protein sequence and structure data, have been used. For structure prediction tasks, especially for fold recognition, statistical models have long been recognized as effective. More recently, protein designs based on a statistical model, the ABACUS (A Backbone-based Amino-aCid Usage Survey) energy function, have also been experimentally verified. To improve current statistical models and to conceive similar approaches for other problems, it is important to understand some of theintriguing issues that affect the accuracy of current statistical energy functions. Liu contributed a mini review on this topic to provide a compact and theoretically-oriented perspective. Besides introducing and discussing the various aspects concerning statistical energy functionsin a related context, it emphasizeson protein sequence design, a topic less covered in the statistical energy function literature. Some distinct issues as compared with structure modeling have been addressed. Resolving them has led the authors to develop the ABACUSenergy function. For details please refer to the article by Haiyan Liu in pages 157－167.

Nov. 2015, Volume 3 Issue 3

Biological investigation is costly and discovery is not necessarily proportional to effort.In sampling experiments such as high throughput sequencing, there is a natural decreasing rate of discovery as a function of sampling effort since with increased sampling it is increasingly more likely to observe a previous discovery.Species accumulation curves are a classical method for modeling the rate of discovery as a function of sampling effort but have previously been limited to small scale experiments. Denget al. present a new method for estimating species accumulation curves and how they can be applied by researchers to explore populations or samples and to optimize experiments to increase biological discovery. With thanks to Dr. Christian Deardorfffor designing the front cover of the issue.

Aug. 2015, Volume 3 Issue 2

Lac repressor, the first discovered transcriptional regulator, is a homodimer protein and presumably would bind to a perfectly symmetric palindromic operator site. However, it was shown that lac repressor binds to its wild-type operator O1 in an intrinsic asymmetric fashion, depending on its central spacer configuration, which makes it behave like a "heterodimer" protein. In this issue, Zuo et al. used Spec-seq approach coupled with site-directed mutagenesis to decipher the underlying mechanism for this unique structural flexibility of DNA binding. Also, a systematic specificity profiling of the whole operator site validates the conventional "additivity assumption" and correlates well with its in vivo occupancy profile.

May. 2015, Volume 3 Issue 1

Microbial communities would serve as the largest reservoir of genes and genetic functions for a vast number of applications in “bio”-related disciplines, including biomedicine, bioenergy, bioremediation, and biodefense. Next-generation sequencing techniques have enabled fast profiling of large volumes of metagenomic samples. As a result, a rapidly increasing number of metagenomic profiles of microbial communities have been archived in public repositories and research labs around the world. Therefore, it is becoming more and more important to perform in-depth analysis for the valuable biological information hidden in large number of samples. 

Mar. 2015, Volume 2 Issue 4

Mining relationships between microbes and the environment they live in is crucial to understand the intrinsic mechanisms that govern cycles of carbon, nitrogen and energy in a microbial community. Building upon next-generation sequencing technology, the selective capture of 16S rRNA genes has enabled the study of co-occurrence patterns of microbial species from the viewpoint of complex networks, yielding successful descriptions of phenomena exhibited in a microbial community. However, since the effects of such environmental factors as temperature or soil conditions on microbes are complex, reliance on the analysis of co-occurrence networks alone cannot elucidate such complicated effects underlying microbial communities. In this issue, Zhu et al. propose a statistical method, which is called Boolean implications for metagenomic studies (BIMS), for extracting Boolean implications to capture the effects of environmental factors on microbial species based on 16S rRNA sequencing data and a Boolean implication network that is constructed between OTUs and environmental factors. Relationships in this network are supported by literature, and, most importantly, they bring biological insights into the effects of environmental factors on microbes. Moreover, the authors apply BIMS to detect three-way relationships and show the possibility of using this strategy to explain more complex relationships within a microbial community.

Jan. 2015, Volume 2 Issue 3

Bacteria in the wild have to face and surmount the challenges raised by fluctuations in extracellular environment. It is observed in a wide range of bacterial species that individual cells within an isogenic bacterial population may stochastically switch among multiple different phenotypes in order to survive in rapidly changing environments. This kind of phenotypic heterogeneity with stochastic phenotype switching is generally understood to be an adaptive bet-hedging strategy. In this issue, Jia et al. propose a nonlinear stochastic model of heterogenous bacterial populations based on a double-positive-feedback gene regulatory network and provide a clear description of phenotypic heterogeneity, stochastic phenotype switching, and bet-hedging within isogenic bacterial populations. Moreover, the authors provide a quantitative characterization of the critical state of heterogenous bacterial populations and develop a data-driven method to identify the critical state without resorting to specific mathematical models. The cover image shows the double-positive-feedback gene regulatory network used in their model and its deterministic and stochastic nonlinear dynamics.

Dec. 2014, Volume 2 Issue 2

The CRISPR-Cas9 system, as a defense mechanism in prokaryotes, has been repurposed as an RNA-guided DNA targeting platform. The CRISPR-Cas9 system as a tool is increasingly used to manipulate the genome and transcriptome. Target specificity of the CRISPR-Cas9 system is a critical issue in applications. In this issue, Xuebing Wu et al. comprehensively described recent progress on the specificity and application of the CRISPR-Cas9 system. The cover image shows a snapshot of the crystal structure of Cas9 (in grey) in complex with a single-guide (sg)RNA (in red) and the target DNA (in green). The target DNA base-pairs with the 5' end of thesgRNA. Formore details please refer to the article by XuebingWu et al. in pages 59—70.

Jun. 2014, Volume 2 Issue 1

Transcriptional regulation plays an important role in regulating many biological processes. Based on genomic information, genome-wide transcriptional regulatory network can be reconstructed by a combined experimental and computational strategy. Biochemical approaches can be used to identify physical occupancies of transcription factors (TFs) on the promoters of genes. Genetic perturbations can provide information on the function of TFs. Transcriptome analysis identifies co-expressed (possibly co-regulated) genes under different conditions. TF-binding sites, which are critical in transcriptional regulation, can be determinedby biochemical experiments or predicted by computation. Many algorithms have been developed to integrate the experimental data to reconstruct the functional dynamic transcriptional regulatory network. (See the article in this issue by Guodong Liu, Antonio Marras and Jens Nielsen)

Sep. 2013, Volume 1 Issue 4

Modeling gene regulatory networks (GRNs) is an important topic in systems biology. Although there has been much work focusing on various specific systems, the generic behavior of GRNs with continuous variables is still elusive. In particular, it is not clear typically how attractors partition among the three types of orbits: steady state, periodic and chaotic, and how the dynamical properties change with network’s topological characteristics. In this issue, Li et al. show that the stability of a random GRN is typically governed by a few embedding motifs of small sizes, and therefore can in general be understood in the context of these short motifs. This finding provides insights for the study and design of genetic networks. The cover image shows a big random network with its small core motifs highlighted in yellow, the 3-dimensional projection of chaotic attractor it may generate in the upper left panel and the 3-dimensional project of a limit cycle attractor in the right bottom panel.

Sep. 2013, Volume 1 Issue 3

Cancer stem cell (CSC) theory suggests a hierarchical structure where CSCs are capable of giving rise to non-stem cancer cells (NSCCs) but not vice versa. However, an alternative scenario of bidirectional interconversions between CSCs and NSCCs was proposed very recently. In this issue, Da Zhou et al. present a general population model of cancer cells by integrating conventional CSC model with direct conversions between different cell states, namely, not only can CSCs differentiate into NSCCs by asymmetric cell division, NSCCs can also dedifferentiate into CSCs by cell state conversion. When applying the model to recent experimental data, it is found that the transient increase in CSCs proportion initiated from the purified NSCCs subpopulation cannot be well predicted by the conventional CSC model where the conversion from NSCCs to CSCs is forbidden, implying that the cell state conversion is required especially for the transient dynamics.

Jun. 2013, Volume 1 Issue 2

The cover image shows a 3D model of the mouse chromosome 6 in mouse embryonic stem cells, predicted by the Bayesian 3D constructor for Hi-C data (BACH). Each sphere represents a topological domain. The volume of each sphere is proportional to the genomic size of the corresponding topological domain. The red, white and blue colors represent topological domains belonging to compartment A, straddle region and compartment B, respectively. Compartment A contains gene rich, actively transcribed, accessible and early replicated chromatin. Compartment B contains gene poor, lowly transcribed, inaccessible and late replicated chromatin. In this 3D model, topological domains with the same compartment label tend to locate on the same side of the structure. The spatial organization of compartment A and B is consistent with their interaction frequencies and the observation that compartment B tends to be associated with the nuclear membrane.

Mar. 2013, Volume 1 Issue 1

Integrated Personalized Omics: Advancements in technology allow for following thousands of molecular components in an individual. It is now possible to obtain whole personal genomes (DNA sequencing), to probe the complete set of expressed genes (RNA sequencing for transcriptomes), profile all the proteins (proteomes) and the levels of all small molecules (metabolomes) in cells, and collect many other omic information. Integrating the individual datasets and incorporating personal medical information allows for the construction of an integrative personal omics profile (iPOP). Multiple complex omics datasets may be visualized in networks, reflecting the dynamic collective behavior of thousands of microscopic components that correspond to the macroscopic physiological state of an individual. An unprecedented level of molecular detail and complexity can be captured and may provide insight to the dynamics of disease onset, progression and treatment, including the possibility of preventive care. The quantitative analysis of such profiles, with emphasis on their dynamic responses to changes in the physiological states of an individual, presents a current possible implementation of longitudinal personalized medicine.