The n_instances is an old value that was a pre-iProphet attempt to take into account the number of observations of a given peptide. It simply computes the sum of probabilities of all spectra matching to a given peptide. The statistic is calculated but in no way affects the protein probability unless you enable the ProteinProphet INSTANCES model, which you don't need to enable when using ProteinProphet on iProphet data.
On Wed, Nov 2, 2011 at 12:21 PM, Gabriel Gray <[email protected]> wrote: > We're a little confused about "n_instances" (and indirectly " > initial_probability" and "nsp_adjusted_probability") in the prot.xml files: > Specifically, if we ran PeptideProphet, iProphet, and ProteinProphet with > MinProbs of 0 (instead of the default of 0.05), would that mean that the > n_instances is including in its calculation the peptide-spectra matches > (PSMs) which originally had peptide probabilities (in the pep.xml file) of > something as low as 0? If so, does that mean that simply because 1 peptide > in one of the pep.xml files had a high probabilty (as indicated by the > "initial_probability" field in the prot.xml file), then all other > occurrences of that same peptide sequence now automatically gets counted in > the prot.xml's "n_instnaces" field, even if the oher occurrences of that > peptide were of extremely low priobability (eg 0.01)? Isn't this inacurate? > That is, should we perhaps have used a value of MINPROB that is not 0 (in > either peptideProphet, iProphet, or ProteinProphet), such as say 90%? > > Or, would it be equally acceptable to bring in all the data (e.g,. MinProb > =0) but then load the results in some sort of relational database and then > dynamically re-calculate n_instances based on the PSM that are NOT filtered > out (e.g., "where PeptideProphet PSM Probability > 90%") or would that > ignore the benefit of the "nsp_adjusted_probability" calculation? > > > -- > You received this message because you are subscribed to the Google Groups > "spctools-discuss" group. > To view this discussion on the web visit > https://groups.google.com/d/msg/spctools-discuss/-/qZKanml8SDgJ. > To post to this group, send email to [email protected]. > To unsubscribe from this group, send email to > [email protected]. > For more options, visit this group at > http://groups.google.com/group/spctools-discuss?hl=en. > -- You received this message because you are subscribed to the Google Groups "spctools-discuss" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/spctools-discuss?hl=en.
