Cancer, according to a 2018 survey, afflicted two million people and killed approximately eight lakhs. In most cases, cancer is undetectable till the last stage. Therefore, simple and easy detectors are in demand for early detection.
Now, Tanusree Roy from the UEM, Kolkata has devised an electro-sensor to distinguish between cancerous and non-cancerous cells based on the length and polarity of protein sequences.
Proteins are made up of amino acids. Amino acids have a common carboxyl and amino group backbone and side chains specific to each amino acid. The sequences determine whether the protein is hydrophobic or hydrophilic.
Tanusree selected forty genes from the NCBI database, of which twenty genes were identified as cancerous and fifteen as non-cancerous. Using MATLAB, the researcher simulated an electrical sensor circuit out of the protein sequences derived from these genes. She designed a circuit with a seven-OHM capacity resistor based on the amino acid backbone: COOH and NH2. To depict the side chains of hydrophilic and hydrophobic amino acids, she selected capacitors and inductors respectively.
She also designed a circuit for an individual amino acid by connecting a specific capacitor or inductor to the resistors in parallel. Tanusree arranged these individual amino acid representatives in-series, to mimic the amino acid chain in biological cells.
When principles from the theory of complex networks were applied, she observed that the circuits showed variations in the phase value for cancerous and noncancerous genes. She deduced that these variations depended on the polarity and charge of the proteins. The proteins of the genes related to cancer have more hydrophilic amino acids than those that are non-cancerous.
The researcher claims that the accuracy of the results achieved by the network-based sensor model is more than 87 percent. The model, she suggests, can help predict cancer related genes.
Gene 685: 62-69 (2019); DOI: 10.1016/j.gene.2018.10.073
Dhanashree, CFTRI, Mysore