Ve·em v1.4
Matthews probability calculator — paste a beamline summary or drop an MTZ file
(ssDNA) or (dsRNA) to a FASTA header to override
Open in Déjà Vu
and probabilities will appear here.
Mission statement & references
- Matthews, B.W. (1968) Solvent content of protein crystals. J. Mol. Biol. 33, 491–497. doi:10.1016/0022-2836(68)90205-2
- Kantardjieff, K.A. & Rupp, B. (2003) Matthews coefficient probabilities: improved estimates for unit cell contents of proteins, DNA, and protein–nucleic acid complex crystals. Protein Sci. 12, 1865–1871. doi:10.1110/ps.0350503
- Weichenberger, C.X. & Rupp, B. (2014) Ten years of probabilistic estimates of biocrystal solvent content: new insights via nonparametric kernel density estimate. Acta Cryst. D70, 1579–1588. doi:10.1107/S1399004714005550
- Weichenberger, C.X. & Rupp, B. (2015) MATTPROB: Matthews coefficient probabilities (developer summary, kernel method, P(n) prior table). Comput. Crystallogr. Newsl. 6, 14–19.
How to use Ve·em
Ve·em estimates how many copies of your molecule fit in the asymmetric unit (ASU) of a crystal, and gives a probability for each candidate based on empirical packing statistics from 60,000+ PDB structures.
Step 1 — supply the unit cell & space group
Two ways:
- Paste any beamline summary, XDS output, HKL2000 log, or CCP4 line directly into the text box. Ve·em extracts the cell and space group automatically and highlights what it found.
- Drop an MTZ file onto the drop zone. The cell, space group, and resolution are read directly from the binary header — no guessing.
Step 2 — enter the molecular weight
Type the MW of the monomer in kDa. For a hetero-complex, enter the combined MW of the whole complex. You can also paste a sequence into the sequence box below — for a complex, paste both chains concatenated or as two separate FASTA entries and the combined MW will be calculated automatically.
Step 3 — read the result
The table lists each candidate copy number with its Matthews coefficient (VM), solvent content, and probability. The most likely copy number is highlighted. The chart shows the empirical solvent distribution with your candidates plotted on it — you can toggle between the VM and solvent % axes.
Optional refinements
- Resolution — narrows the reference dataset to structures of similar or better resolution, giving a sharper probability estimate.
- Sample type — switch to nucleic acid or complex if appropriate (datasets are smaller, so treat those probabilities with more caution).
- Multimer unit — if you know the biochemical oligomeric state (e.g. a dimer), set this so candidates are multiples of that number.
- P(n) prior — optionally weights candidates by how common each copy number is across the PDB.
Tips
- All fields are editable after auto-fill — just type to override.
- Results update live as you type; no submit button needed.
- A VM between 1.7 and 4.0 ų/Da is physically reasonable; the distribution peaks near 2.15 ų/Da (~43 % solvent).