No abstract

No abstract

Bioinformatics-based applications have been incorporated into several medical disciplines, including cancer, neuroscience, and recently psychiatry. Both the increasing interest in the molecular aspect of neuropsychiatry and the availability of high-throughput discovery and analysis tools have encouraged the incorporation of bioinformatics and neurosystems biology techniques into psychiatry and neuroscience research. As applied to neuropsychiatry, systems biology involves the acquisition and processing of high-throughput datasets to infer new information. A major component in bioinformatics output is pathway analysis that provides an insight into and prediction of possible underlying pathogenic processes which may help understand disease pathogenesis. In addition, this analysis serves as a tool to identify potential biomarkers implicated in these disorders. In this chapter, we summarize the different tools and algorithms used in pathway analysis along with their applications to the different layers of molecular investigations, from genomics to proteomics.

The crucial biological role of proteases has been visible with the development of degradomics discipline involved in the determination of the proteases/substrates resulting in breakdown-products (BDPs) that can be utilized as putative biomarkers associated with different biological-clinical significance. In the field of cancer biology, matrix metalloproteinases (MMPs) have shown to result in MMPsgenerated protein BDPs that are indicative of malignant growth in cancer, while in the field of neural injury, calpain-2 and caspase-3 proteases generate BDPs fragments that are indicative of different neural cell death mechanisms in different injury scenarios. Advanced proteomic techniques have shown a remarkable progress in identifying these BDPs experimentally. In this work, we present a bioinformatics-based prediction method that identifies protease-associated BDPs with high precision and efficiency. The method utilizes state-of-the-art sequence matching and alignment algorithms. It starts by locating consensus sequence occurrences and their variants in any set of protein substrates, generating all fragments resulting from cleavage. The complexity exists in space O(mn) as well as in O(Nmn) time, where N, m, and n are the number of protein sequences, length of the consensus sequence, and length per protein sequence, respectively. Finally, the proposed methodology is validated against βII-spectrin protein, a brain injury validated biomarker.Degradomics discipline has been recently introduced to depict the application of an omics approach (genomics and proteomics etc.) to identify different proteases and their subsequent proteolytic substrates/degradome in a defined pathophysiological condition 1 . Recently, the use of bioinformatics tools as means for data mining has spanned different fields in cancer, neuroscience and biochemistry research 2,3 . Degradomics as a discipline has benefitted from data mining strategies as tools to predict degradome specific substrates in silico [4][5][6][7] . However, the application of bioinformatic tools on degradomics analysis requires different types of sequencing matching algorithms making it one of the challenging fields despite its potential beneficial outcomes mainly in clinical and diagnostic research. Knuth et al. developed an algorithm that only finds exact matches of a subsequence of size m in a sequence of size n in O(m + n) 8 . It is worth to know that other algorithms have identified sequence variants with comparable complexity, but not with the same fidelity. Lipman et al. devised a heuristic algorithm called FAST Protein (FASTP) 9 ; it is based on alignment approach and is both rapid and sensitive in finding similarities between any amino acid subsequence and matching sequences in a database. Yet, it does not cover all regions, as it starts with an anchoring scheme that identifies identical regions using a replaceability matrix 9 . Similarly, Altschul et al. developed another heuristic algorithm BLAST, along with its variations; this algorithm supposedly supersedes FASTP ...

No abstract