TPP and NPA analysis

The TPP/NPA analysis module is designed for classic CETSA experiment for target deconvolution. ProSAP could accepts more or less temperature points but will require minimum six temperature points for inferring melting temperature (Tm) of protein. In this module, users can identify potential targets with parametric TPP analysis or nonparametric analysis (NPA). User can input the preprocessed data file of case and control group, with or without replicates.

Input

Data should be upload via Data menu, and click Load proteomics. You can upload multiple files in the same time. The data table can be the output of the preprocessing procedure. If not, please make sure the format is consistent to the example files. The data table should also include Accession column, but do not select this column in the selection dialog. Only choose the columns including protein abundances is OK. Then, please appoint each file as case sample #1, control sample #1, case sample #2, control sample #2. The last two of them are not compulsory. However, experimental replicates are highly recommended for reducing the false positive.

After input, you can check and adjust the parameters with clicking the Params button. Then, select a statistical method for preprocessing the uploaded data with the Analysis menu. There are three methods for choosing: TPP method, NPARC method and Distance based method.

Each data table of the input should include the following columns:

  • Accession: This column is necessary, which is the ID of the protein. It can be uniprot id or/plus gene name or anything else.
  • Tx: These columns should include the relative abundances of each protein. The abundances are normalized by dividing the abundances under 37 C. The x in the column name should be replaced with the temperture, which can be integer or float.

If there are values much higher than 1 in your input table, there will be a warning reminding you that the data may be un-normlized. If you are sure that you did, ignore it.

Parameters

TPP method

The workflow of TPP analysis is implemented as described in the Savitski protocol. See the reference for the details. The parameters includes:

  • Min R2: The threshold of the goodness-of-fit of the melting culve fitting. Default: 0.8
  • H-Axis: The threshold of how much protein is precipitated for calculating the Tm. Default: 0.5
  • Max Plateau: Plateau (lower horizontal asymptote) of the fitted melting curve under the given condition. Default: 0.3
  • Replicate Check: Whether to use the filtering criteria based on the replicates:
    (1) One of the P values for the two replicate experiments < p value threshold #1 and the other one p value threshold #2
    (2) The coefficient of determination R2 are over the user-defined thresholds.
    (3) The minimum slope in each of the control versus treatment experiments is lower than user-defined thresholds.
    (4) The melting point shifts in the vehicle versus treatment experiments have the same direction.
    (5) Both melting point differences in the two pairs of control versus treatment experiments are greater than the melting point difference between the two vehicle controls.
  • Min Slope: Slope threshold of the fitted melting curve at the inflection point under the given condition

NPARC method

The workflow of NPARC analysis is implemented as described in the Dorothee Childs's paper. See the reference for the details. The parameters includes:

  • Min Null R2: The threshold of the goodness-of-fit of the Null model. Default: 0.8
  • Min Alt R2: The threshold of the goodness-of-fit of the Alternative model. Default: 0.8
  • Max Plateau: Plateau (lower horizontal asymptote) of the fitted melting curve under the given condition. Default: 0.3

Distance method

Distance based model is more straightforward. The relative fold changes of each protein of case and control groups are fitted into melting curves, respectively. The distance of the protein between case and control group is calculated for evaluating the changing of the thermal stability. The parameters includes:

  • Metrics: Which function is used for calculating the distance.
  • Min R2: The threshold of the goodness-of-fit of the melting culve fitting. Default: 0.8
  • Max Plateau: Plateau (lower horizontal asymptote) of the fitted melting curve under the given condition. Default: 0.3

Display

  • Replicate #1 Protein Table: The quantitative table of the protein groups of the uploaded Replicate #1.
  • Replicate #2 Protein Table: The quantitative table of the protein groups of the uploaded Replicate #2.
  • Result table: The output of the preprocessing results. Click Save to write to csv file. The explaination of the columns:
    • Score: The harmonized score combining both the significance (p-value) and the goodness-of-fit (R2).
    • Rep1pVal (-log10): -log10 p-value of the ΔTm (TPP) or other distance metrics (NPA) of the Replicate #1.
    • Rep1delta_Tm: Calculated ΔTm of the the Replicate #1 (TPP only).
    • Rep1Group1_R2: Goodness-of-fit of the melting curve of the first group (control) of the Replicate #1.
    • Rep1Group2_R2: Goodness-of-fit of the melting curve of the second group (case) of the Replicate #1.
    • Rep1Group1_Tm: Calculated Tm of the melting curve of the first group (control) of the Replicate #1 (TPP only).
    • Rep1Group2_Tm: Calculated Tm of the melting curve of the second group (case) of the Replicate #1 (TPP only).
    • Rep1min_Slope: The minimum slope of the fitted melting curve of the Replicate #1 (TPP only).
    • Rep2pVal (-log10): -log10 p-value of the ΔTm (TPP) or other distance metrics (NPA) of the Replicate #1.
    • Rep2delta_Tm: Calculated ΔTm of the the Replicate #2 (TPP only).
    • Rep2Group1_R2: Goodness-of-fit of the melting curve of the first group (control) of the Replicate #2.
    • Rep2Group2_R2: Goodness-of-fit of the melting curve of the second group (case) of the Replicate #2.
    • Rep2Group1_Tm: Calculated Tm of the melting curve of the first group (control) of the Replicate #2 (TPP only).
    • Rep2Group2_Tm: Calculated Tm of the melting curve of the second group (case) of the Replicate #2 (TPP only).
    • Rep2min_Slope: The minimum slope of the fitted melting curve of the Replicate #2 (TPP only).
  • Melting curve: Display the fitted melting curve and the original data points.