Bioinformatics
From sequence analysis to multiomics data, we can analyze anything with the right coding and approach
Metagenomics of microbiome during Euglena fermentation: 2021
In this study we explored the microbiome composition during fermentation of Euglena. The objective of this study is to understand the relationship between microbiome and contaminants produced during fermentation like histamines and VBNs.
Cyanobacterial protein physico chemical property database
During protein expression, several proteins often form inclusion bodies due to unsuitable buffer environments. This is due to the specific physicochemical properties that are governing the active state of the protein. Understanding these properties could help us purify a protein in its active state. To assist Cyanobacterial researchers in such research, we developed a database of all properties by analyzing the physiochemical properties and their inclusion body formation confidence rate.
​
Representation of Lipidome data as a network
In this study, I designed a PERL & PYTHON based algorithm to sort the lipidome data into nodes representing lipid class, fatty acids or carbon:double bond information. A piple line architecture was also designed to create an input file for BIOLAYOUT3D software.
For more information and code: https://tsukuba.repo.nii.ac.jp/records/47391
Pattern matcher for gene or protein sequences
Researchers are often interested to identify specific patterns of gene segment or SNPs or Protein motifs like metal ion binding domain or di/tri peptide patterns in the respective sequences. However, it is very difficult to identify identical patterns or similar patterns or patterns with random chance sites (NNN) manually or even using available online tools. Here, I developed a PERL based pattern recognition algorithm, that can do these things. If you are interested you can directly contact me for the analysis.
Metalloprotein identification tool
Presence of cofactors/metal ions in protein assist in their functional activity. Identification of a proteins specific metal binding characteristic could help annotate several hypothetical proteins of Cyanobacteria. Identification of such feature through crystallography is cost effective, therefore we employed sequence based analysis by application of neural networks and HMM modeling to predict specific metalbinding sites proteins.