Team:Shanghai-United/Model

The characterization protein and early diagnosis of cervical cancer

Here we demonstrate specific protein expression variations related to NFX1 over-expression, a gene found to be closely associated with cervical cancer.

Model

Model Part 1 Protein Assay

After IP and FASP, we loaded both groups of our samples on Mass Spectrometer. The mass spectrometry was operated to get statistics of proteins in each sample using electric and magnetic fields. We chose to analysis our data of LFQ intensity, for it is normalized to exclude outliers to best represent the ratio changes of different samples.

Here are the detailed steps of how we approached the data (operated in excel)

1. First filter out all proteins with a mark in reverse and potential contaminant column. This helps exclude experimental error caused by contamination.
2. Out of the filter proteins, copy their protein names, gene names, and LFQ intensities for all 6 samples (3EV and 3NFX1) to create a new sheet.
3. Calculate and filter out any group of 3 (EV/NFX1 separately) that has less than one value out of three. All proteins left after this step should have at least 2 values in both groups’ intensity (EV and NFX1)
4. Substitute all values of 0 for NA, so it could be read and run in R studio.
5. Write the codes for k-nearest neighbor (KNN) algorithm to estimate the missing value NA.
6. Get the MOM (multiple of median) value for all values: calculate median for each group of three and divide each individual value by the median of its group, put the result into a new sheet.
7. Calculate the average of each group of three for every protein, then divide the average of experimental group by control group to get the fold change.
8. Apply the TTest function in excel (2 tails, type 2) to get the p value.
9. Filter to get the proteins with a fold change bigger than or equal to 1.2 or less than or equal to 0.8, and those with a p value less than or equal to 0.05. Those proteins have a significant change in concentration when NFX1 is overexpressed, implying their strong relationship with the gene that is viral to HPV infection.
10. With the final data, we can create several graphs to visualize the results.

Model Part 2

During our two experiment, FASP and IP, we used Tryptophan and BCA quantification to determine the concentration of protein in our samples. We acquired standard Tryptophan and BCA sample, and used their results to generate a formula to interpret data from our own samples. Below are the data we collected and how we approached them.

©2019 wiki-hpv Corporation. All rights reserved.