User guide:
1. The proteomics data for 347 patient-derived xenograft (PDX) models across 23 cancer types was generated using 4D-DIA proteomics technology, a label-free approach that quantifies relative protein expression levels among samples. Raw data was processed by Spectronaut v17, during which protein expression values were quantified with MaxLFQ algorithm and cross sample normalization was performed with Local Normalization method. Log2 transformation was applied to the resulting values from Spectronaut for protein expression quantification. The protein expression profiles of select models were depicted using a barplot, while a Scatter plot featuring a regression line was employed to elucidate the correlation between protein expression data and corresponding gene expression data.

2. To reduce the interference of human-mouse common peptides in proteomics profiling of PDX models, we performed mass spectrometry data acquisition after removing mouse stromal cells (separate-then-run). These peptides are abundant in mouse stromal cells and can affect human protein quantification. More details about our separate-then-run process are described in our research article [1].

3. Users can access the expression of a target protein through searching with the gene symbol encoding it. The results are shown in a table below the barplot, including protein groups (PGs), PG expression values, and molecular weights (in Daltons) of the proteins in each PG. A protein group refers to protein(s) identified with the same set of peptides. Each protein group can either has a single protein or multiple proteins (all proteins are represented in UniProt accession numbers and separated by semicolons). For example, if a user searches for protein expression of the gene “ATP5F1E”, they will get a protein group containing two proteins (P56381 and Q5VTU8). The multiple proteins in a protein group commonly have high sequence similarities so that their exact expression values cannot be clearly separated due to common peptides. In this database, >99% protein groups have only a single protein.

4. To evaluate a protein's expression level relative to all proteins, users can simply click on the gene's name in the Protein Expression table. Subsequently, a Comparative Distribution figure illustrating Proteomic and Transcriptomic Expression Levels will automatically load below the Scatter plot. This image provides a visual representation of the selected protein and its corresponding gene within the overall expression landscape, aiding in the determination of whether the target protein exhibits high or low expression.

5. Occasionally, different models may have different protein groups assigned to the same protein. This is due to experimental variations (e.g., models were measured in different batches) or gene-isoform relationships. Therefore, users should avoid comparing models with different protein groups, as they are not comparable. For example, through querying gene “LGALS9B”, users will get two different protein groups: one is “Q3B8N2” and the other is “Q3B8N2;Q6DKI2”. The values of the two different groups are not comparable, and the barplot will group and color them differently for easy visual comparison.

Reference:
[1] Zhaomei Shi, Binchen Mao, Xiaobo Chen, Piliang Hao, Sheng Guo; Mouse Stromal Cells Confound Proteomic Characterization and Quantification of Xenograft Models. Cancer Research Communications 1 February 2023; 3 (2): 202–214. https://doi.org/10.1158/2767-9764.CRC-22-0431

Disclaimer:
CROWN BIOSCIENCE DOES NOT MAKES ANY REPRESENTATIONS OR WARRANTIES OF ANY KIND, EITHER EXPRESS OR IMPLIED, AND EXPRESSLY DISCLAIMS ALL IMPLIED WARRANTIES OF TITLE, NON-INFRINGMENT, MERCHANTIBILITY, AND FITNESS FOR A PARTICULAR PURPOSE OF THIS DATABASE.