Current Bioinformatics - Volume 3, Issue 2, 2008

Proteins localize to many intracellular and extracellular organelles in the malarial parasite Plasmodium falciparum. Although organellar localization in the parasite has been studied in detail, there is a need for a comprehensive ruleset that captures the different paths and mechanisms used by proteins for localization. This review is an attempt to build such a rule-set through a combination of rules gleaned from literature reports. A prototype localization prediction tool has also been implemented by incorporating certain sequence/composition rules from the rule-set.

The prediction of protein secondary structures is not only of great importance for many biological applications but also regarded as an important stepping stone for solving the mystery of how amino acid sequences fold into tertiary structures. Recent research on secondary structure prediction is mainly based on widely known machine learning techniques, such as Artificial Neural Networks and Support Vector Machines. The most significant breakthroughs were the incorporation of new biological information into an efficient prediction model and the development of new models which can efficiently exploit suitable information from its primary sequence. Hence this paper reviews the theoretical and experimental literature of these models with a focus on informational issues involving evolutionary and long-range information of protein sequences. Furthermore, we investigate several key issues in protein data processing which involve dimensionality reduction and encoding schemes.

In this era of whole genome sequencing, reliable genome annotations (identification of functional regions) are the cornerstones for many subsequent analyses. Not only is careful annotation important for studying the gene and gene family content of a genome and its host, but also for wide-scale transcriptome and proteome analyses attempting to describe a certain biological process or to get a global picture of a cell's behavior. Although the number of sequenced genomes is increasing thanks to the application of new technologies, genome-wide analyses will critically depend on the quality of the genome annotations. However, the annotation process is more complicated in the plant field than in the animal field because of the limited funding that leads to much fewer experimental data and less annotation expertise. This situation calls for highly automated annotation platforms that can make the best use of all available data, experimental or not. We discuss how the gene prediction (the process of predicting protein gene structures in genomic sequences) research field increasingly shifts from methods that typically exploited one or two types of data to more integrative approaches that simultaneously deal with various experimental, statistical, or other in silico evidence. We illustrate the importance of integrative approaches for producing high-quality automatic annotations of genomes of plants and algae as well as of fungi that live in close association with plants using the platform EuGène as an example.

There is a flood of molecular data about many aspects of cellular functioning. This data ranges from sequence and structural data to gene and protein regulation data, including time dependent changes in the concentration. Integration of the different datasets through computational methods is required to extract biological information that is relevant from a systems biology perspective. In this paper we discuss how different computational tools and methods can be made to work together integrating different types of data, mining these data for biological information, and assisting in pathway reconstruction and biological hypotheses generation. We review the recent body of literature where such integrative approaches are used and discuss automation of data integration and model building to generate testable biological hypotheses. We analyze issues regarding the design of such automated tools and discuss what limitations and pitfalls can be foreseen for the automation and what solutions can computer science and biologists provide to overcome them.

Synthetic biology is an emerging field that strives to build increasingly complex biological networks through the integration of molecular biology and engineering. The growth of the field has been supported by progress in the design and construction of synthetic genetic and protein networks. This has led to the possibility of assembling modular components to attain novel biological functions and tools. In addition, these synthetic networks give rise to insights that facilitate the investigation of interactions and phenomena in naturally-occurring networks. Integration of well-characterized biological components into higher order networks requires computational modeling approaches to rationally construct systems that are directed towards a desired outcome. A computational approach would improve the predictability of the underlying mechanisms that would otherwise be difficult to deduce through experimentation alone. The analysis and interpretation of both qualitative and quantitative models also becomes increasingly important towards taking a systems-level perspective on synthetic genetic and protein networks. This review will first discuss the analogy of synthetic networks to circuit engineering. It will then look at computational modeling approaches that can be applied to biological systems and how synthetic biology will help to develop more accurate in silico representations of these systems.