CMRI's ProCan team recently published a new method for removing the noise in data, which will not only aid ProCan's efforts to analyse data from tens of thousands of cancer samples but other MS-based proteomics research internationally. The research was published in Bioinformatics and led by Dr Qing Zhong and the ProCan Data Science Group.
They used advanced computational methods to analyse raw mass spectrometry (MS)-based proteomic data. Their paper outlines a current shortcoming in the data pre-processing process and presents a novel technique that addresses this limitation.
MS-based proteomics is an analytical approach used to identify and quantify proteins and peptides, which are the building blocks of the sample. While this technique is incredibly advanced, it still creates a small amount of digital noise that muddles and disturbs the data, making it less reliable. While there are tools currently available for denoising MS-based proteomic data, they cannot meet the necessary accuracy and throughput for ProCan’s aims. Dr Zhong and the ProCan team at CMRI have developed a brand-new method for denoising MS data, which outperforms the competing methods.
“Noise has a negative impact on the identification and quantification of peptides, which influences the reliability of the MS- based proteomic data. We have developed a computational method to reduce the noise on MS data by treating them as two-dimensional images. This strategy has enabled us to use an advanced image analysis algorithm (undecimated wavelet transform) to transform MS data to the wavelet domain prior to denoising. “