For low through put, use ExPaSy:

http://expasy.org/proteomics/post-translational_modification

Thanks Qing-Rong for the answer!

I would like to focus more on high troughput.

which type of analysis you want to do? I have done and already doing such type of work. If you go through such publications containing "interplay between phosphorylation/ glycosylation" etc u can find many papers, and u can also compare the results of different servers predicting PTMs.

I think all methods are going to be highly CPU intensive. You increase the search space and it's going to take time.

I haven't used it, but DeltaAMT (10.1074/mcp.M110.000455 ) says it's fast. MODa and MODi (10.1074/mcp.M111.010199, 10.1021/pr9001146) are worth a look, but I've found the software from this lab quite buggy and hard to get working. PTMFinder is yet another one (10.1021/pr070444v), but I haven't tested this either. That's all I have in my list that you haven' t already mentioned.

Unfortunately, I find a lot of the scientific programs buggy and just refuse to work on any system except the authors! Fortunately, there are many examples of good free software available to us. I'd be interested to hear how you go with the various programs.

Peter

JUST LOOK ON THESE PAPERS

WE MOSTLY USED SOFT WARES PROVIDED ON EXPASY WEBSITE

Ahmad W, Shabbiri K, Ijaz B, Asad S, Nazar N, et al. (2011) Serine 204 phosphorylation and O-beta-GlcNAC interplay of IGFBP-6 as therapeutic indicator to regulate IGF-II functions in viral mediated hepatocellular carcinoma. Virol J 8: 208.

Ahmad W, Shabbiri K, Ijaz B, Asad S, Sarwar MT, et al. (2011) Claudin-1 required for HCV virus entry has high potential for phosphorylation and O-glycosylation. Virol J 8: 229.

Ahmad W, Shabbiri K, Nazar N, Nazar S, Qaiser S, et al. (2011) Human linker histones: interplay between phosphorylation and O-beta-GlcNAc to mediate chromatin structural modifications. Cell Div 6: 15.

Kaleem A, Hoessli DC, Haq IU, Walker-Nasir E, Butt A, Iqbal Z, Zamani Z, Shakoori AR; Nasir-ud-Din: CREB in long-term potentiation in hippocampus: role of post-translational modifications-studies In silico. J Cell Biochem 2011, 112:138-146.

Ahmad I, Khan TS, Hoessli DC, Walker-Nasir E, Kaleem A, Shakoori AR, Nasir-ud-Din: In silico modulation of HMGN-1 binding to histones and gene expression by interplay of phosphorylation and O-β-GlcNAc modification. Protein Pept Lett 2008, 15:193-199.

Kaleem A, Hoessli DC, Ahmad I, Walker-Nasir E, Nasim A, Shakoori AR, Nasir-ud-Din: Immediate-early gene regulation by interplay between different post-translational modifications on human histone H3. J Cell Biochem 2008, 103:835-851.

Our minimotif miner tool at minimotifminer.org has verified and predicted PTMs for most types

Hi, I tested a lot and I run PEAKS now, it is slow on my computer but I think a faster one can make the job better. I look at PTMs since several years and have seen many nonsense (also published) from many approaches. There are simple ones which immediately can be foudn and one which would require manual checking, nothing for any throughput. We did it for several thousands of phosphosites (I mean manual checking) and found for MASCOT that below a certain threshold it is no sense even to look at the MS/MS spectra, then there is a range where may be some believable ones and a range above where most are correct. To make it more complicated if the phosphorylation is close to the N-terminus of the peptide you may end up with a neutral loss and no localization of the site.

The software are improving and new papers recently showing also new approaches, hope they make it soon into our computer.

phosphositeplus, Netphos, Group based prediction system (GPS2)

Many thanks for your answers.

A lot of tool that you suggested (like the ones in Expasy or Netphos and GPS2) are specific for one modification. What I was looking for are tools (like Peter's ones) that with one analysis can find all the possible modifications of, for example, an entire proteome.

The minimotif commented by Martin is a nice tool but you can work with only one protein and it is an in silico prediction. I would like comments on tools that works with real data.

I am trying to run PeaksPTM with all the possible unimod modification but I think that the computer will collapse very soon. InspecT and SIMS are very complicated to install and take 3 days of computation per sample, plus the time to parse the results. pMatch and PilotPTM do not work for me.