This Document provides an overview of the methods used for PDX proteomic and subsequent data analysis. The primary objective is to identify proteins that are secreted in substantial quantities and those that are differentially abundant in TP53 mutant as compared to nonmutant PDX tumors. These samples are taken from patient PDX lines who exhibit unique genetic variants in the source of the Interstitial Fluid (IF) and Tumor Lysate (TL).
Each patient's samples undergo a meticulous process that includes cutting a specific amount of tissue, washing, blotting, and incubating it before it is subjected to various levels of centrifugation. The proteins in these samples are then analyzed, with an emphasis on identifying two main types of proteins: those that are heavily secreted, and those that are differentially abundant in TP53 mutant compared to nonmutant PDX tumors.
Deposited summary table of proteins identified through
Location: server Gade Lab/Projects/PDX_TIF_proteomics_2021/TP53_analysis_Oct2021
The analyses primarily focused on comparing proteins in TP53 mutant (mut) vs nonmutant (nonmut) samples within both Interstitial Fluid (IF) and Tumor Lysate (TL), and the correlation of these proteins with RNAseq data.
The analyses use the DEP R package, which was chosen for its clean functions and capabilities in performing differential expression analysis, as well as its built-in tools for filtering and imputation. A careful imputation process was apply in the datasets due to a common issue in proteomic.
The first round of analysis identified 22 significant proteins in the group comparison between TP53 mut vs. non TP53 mut of TIF. This was achieved using a mixed imputation approach, distinguishing between "Missing at Random" (MAR) and "Missing not at Random" (MNAR) values. The same settings led to the identification of 58 significantly regulated proteins in the TL sample set.
The subsequent analyses involved the creation of a gene list using parameter comparison(1_mixed_MLE_zero_1 ) and correlation between IF and TL. The aim was to maximize the number of genes identified and use the gene list for enrichment analysis and cross-comparison with other datasets.
The analysis also identified proteins that had higher Label-Free Quantification (LFQ) values in IF samples versus TL samples. This was test through paired t-tests, which identified 98 proteins when using non-adjusted p-values. And 141 genes when repeated using a one-tailed t-test.