3.4.2 Statistical Analysis
Using Perseus
1. Download Perseus from https://www.maxquant.org/per
seus/ website.
2. Load proteinGroups.txt file into Perseus using the generic
upload button (Green arrow at the top left of the Perseus
window) (see Note 35).
3. Select LFQ intensities as expression data by transferring them
in the main box and iBAQ intensities by transferring them in
the numerical box for and press “load” (see Note 36).
4. The matrix 1 open containing all the data necessary for the
statistical analysis. First filter the dataset using “Filter rows/
Filter based on categorical column” and remove rows positive
to “identified by site,” “reverse,” and “potential contaminant”
(see Note 37).
5. Transform intensities in log2 values using “Basic/Transform.”
6. Annotate the data set according to the conditions groups using
“Annot. Rows/Categorical annotation rows”, for example two
groups Control and Pulldown (see Note 38).
7. Filter the dataset to eliminate proteins quantified in only few
conditions using “Filter rows/Filter rows based on valid
values.” It is a best practice to use the default values “Percentage 70” in at least one group (see Note 39).
8. It is a good practice to check the correlation between column
using “Basic/Column correlations” and then applied the
“Hierarchical clustering” without clustering calculation (deselect “Rows tree” and “Columns tree” buttons). Press the multicolored wheel to see the correlation scale.
9. Replace missing LFQ values using “Imputation/Replace missing values from normal distribution” using default parameters.
(Note: Default parameters are mainly useful; however, you can
modify the width and Down shift to impute values for example
at lower intensities.)
10. Check imputation using “Histogram” button and in the histogram window select “selection from imputation” to see in red
imputed values on the blue distribution.
11. Check the relative enrichment between the two groups (e.g.,
Control versus Pulldown) using the “Volcano plot” button.
Select as the first group the Pulldown group and as the second
group the Control group. Press “OK.” Keep others parameters
by default. You can play with the FDR and the s0 values to
check different settings. When you are satisfied of your settings, save the Volcano plot using the “PDF” button (Fig. 4b)
(see Notes 40 and 41).
12. Calculate the relative protein enrichment between the two
groups using “Test/Two samples tests.” Select the same first
and second groups as previously (e.g., respectively Control and
Identification of Associated PDZ Proteins
33
Using Perseus
1. Download Perseus from https://www.maxquant.org/per
seus/ website.
2. Load proteinGroups.txt file into Perseus using the generic
upload button (Green arrow at the top left of the Perseus
window) (see Note 35).
3. Select LFQ intensities as expression data by transferring them
in the main box and iBAQ intensities by transferring them in
the numerical box for and press “load” (see Note 36).
4. The matrix 1 open containing all the data necessary for the
statistical analysis. First filter the dataset using “Filter rows/
Filter based on categorical column” and remove rows positive
to “identified by site,” “reverse,” and “potential contaminant”
(see Note 37).
5. Transform intensities in log2 values using “Basic/Transform.”
6. Annotate the data set according to the conditions groups using
“Annot. Rows/Categorical annotation rows”, for example two
groups Control and Pulldown (see Note 38).
7. Filter the dataset to eliminate proteins quantified in only few
conditions using “Filter rows/Filter rows based on valid
values.” It is a best practice to use the default values “Percentage 70” in at least one group (see Note 39).
8. It is a good practice to check the correlation between column
using “Basic/Column correlations” and then applied the
“Hierarchical clustering” without clustering calculation (deselect “Rows tree” and “Columns tree” buttons). Press the multicolored wheel to see the correlation scale.
9. Replace missing LFQ values using “Imputation/Replace missing values from normal distribution” using default parameters.
(Note: Default parameters are mainly useful; however, you can
modify the width and Down shift to impute values for example
at lower intensities.)
10. Check imputation using “Histogram” button and in the histogram window select “selection from imputation” to see in red
imputed values on the blue distribution.
11. Check the relative enrichment between the two groups (e.g.,
Control versus Pulldown) using the “Volcano plot” button.
Select as the first group the Pulldown group and as the second
group the Control group. Press “OK.” Keep others parameters
by default. You can play with the FDR and the s0 values to
check different settings. When you are satisfied of your settings, save the Volcano plot using the “PDF” button (Fig. 4b)
(see Notes 40 and 41).
12. Calculate the relative protein enrichment between the two
groups using “Test/Two samples tests.” Select the same first
and second groups as previously (e.g., respectively Control and
Identification of Associated PDZ Proteins
33