LD FAQ
Q: Where is the LongDistances software available for download?
A: The newest version can be downloaded from the Hubbell Lab Website.
Q: Why that name?
A: Many years ago, a simple program to estimate inter-spin distances from dipolar broadening was released and called ShortDistances. As the name implies, it was only useful for relatively short distances (<25A). After we obtained pulse instrumentation and started doing DEER, much longer distances could be measured accurately. It was only natural to call it LongDistances for symmetry. The name intentionally does not include terms such as DEER, PDS, or Peldor to remain neutral in the current naming heterogeneity of the method.
Q: Are there any instructions or tutorials?
Online help is available.
Q: Why don't you use SI units?
A: We are aware that customary units show regional variations. While the US still clings to e.g. Angstroms and Gauss, the rest of the world has adapted SI units such as nm and Tesla. LongDistances currently uses Angstroms for Distance units. If there is sufficient demand, user selectable units might be implemented in the future. Still, I expect that the typical user is sufficiently skilled to correctly apply a factor of ten if needed.
Q: What are some limitations of the program?
A: The program is designed to analyze pairwise interspin distances in the presence of a background of randomly arranged other radicals, such as typically encountered in doubly-labeled dilute proteins (or homodimers of singly labeled proteins). A rich selection of background models is available. If the data is poor (too short, too noisy), no algorithm can get reliable results. Garbage in, garbage out!
The program is not designed to analyze data for more complicated scenarios (This list is probably incomplete)
- One example would be encountered if more than two spins interact. However, if the modulation depth is relatively shallow. a good estimate can still be obtained.
- Another problem is orientation selection. In the presence of strong orientation selection the fit might not be great, but often the distance distribution is still approximately correct.
- Relative peak areas could show distortions if components of the same sample differ dramatically in phase memory time (Baber, J. L., Louis, J. M. and Clore, G. M. (2015),. Angew. Chem. Int. Ed., 54: 5336–5339).
If there are doubts, users are encouraged to compare results with DeerAnalysis By Gunnar Jeschke and get a second opinion. It can be downloaded freely from the EPR@ETH software site, but requires a valid MATLAB license to run.
Q: I wrote a paper and used LongDistances. How should I reference it?
A: There is currently no publication dedicated to the program, just this web site. A reference should mention the words "LongDistances" and "Altenbach" and should point to the software page of the Hubbell lab website hosted at UCLA. It will have download links as well as links to the full product page here.
Q: Is anyone actually using this?
A: There is a significant user-base and I held personal workshops at many locations (e.g. Milwaukee, WI, Santa Barbara, CA, Columbia, MO).
LongDistances is not spyware so I have no download or usage statistics except for private questions initiated by users and submitted bug reports. A good estimate can be gauged by publications that mention the program. Early versions of the program existed long before it was first made public in 2014. The (incomplete) list below will be split into two sections accordingly.
--- NOTE: Any addition and correction to the tables is very welcome! ---
Publications that used private versions:
Christian Altenbach, Ana Karin Kusnetzow, Oliver P. Ernst, Klaus Peter Hofmann, and Wayne L. Hubbell, High-resolution distance mapping in rhodopsin reveals the pattern of helix movement due to activation. PNAS,105:7439-7444 (2008). (link)
López et al, Conformational selection and adaptation to ligand binding in T4 lysozyme cavity mutants, PNAS 2013. (link)
... list is incomplete ...
Publications that mention the public program:
- Stein et al, A straightforward approach to the analysis of Double Electron-Electron Resonance Data, in: Electron Paramagnetic Resonance Investigations of Biological Systems by Using Spin Labels, Spin Probes, and Intrinsic Metal Ions Part A, (Qin et al, ed.) Methods in Enzymology 563:531-567, 2015 (link)
- Thompson et al, Bifunctional Spin Labeling of Muscle Proteins: Accurate Rotational Dynamics, Orientation, and Distance by EPR, in: Electron Paramagnetic Resonance Investigations of Biological Systems by Using Spin Labels, Spin Probes, and Intrinsic Metal Ions Part B, (Qin et al, ed.) Methods in Enzymology, 564:103-123, 2015. (link)
- Gerogieva, Nanoscale lipid membrane mimetics in spin-labeling and electron paramagnetic resonance spectroscopy studies of protein structure and function, Nanotechnology Review 6 (2017) (link)
- Schultz & Klug, High-pressure EPR spectroscopy studies of the E. coli lipopolysaccharide transport proteins LptA and LptC, Appl Magn Reson, 2017 (link)
- Gu et al, Conformational heterogeneity of the allosteric drug and metabolite (ADaM) site in AMP-activated protein kinase (AMPK), Journal of Biological Chemistry, 293:16994-17007, 2018 (link)
- Stadtmueller et al, DEER Spectroscopy Measurements Reveal Multiple Conformations of HIV-1 SOSIP Envelopes that Show Similarities with Envelopes on Native Virions, Immunity, 49:235-246, 2018, (link)
- WIngler et al, Angiotensin Analogs with Divergent Bias Stabilize Distinct Receptor Conformations, Cell, 176: 468-478, 2019. (link)
- Nilaweera et al. Disulfide Chaperone Knockouts Enable In Vivo Double Spin Labeling of an Outer Membrane Transporter, Biophysical Journal, 117:1476-1484, 2019. (link)
- ... list is incomplete ...
Q: Is there a list of programs for the analysis of DEER or Peldor data in terms of distance distributions?
A: Yes! Others have also written their own analysis programs. If it is available to the public, please contact me so it can be included in the list below.
--- NOTE: Any addition and correction to the table is very welcome! ---
- LongDistances (Christian Altenbach, UCLA)
- A full featured program for typical data. Under active development with new features added regularly
- Advantages: Easy to learn and use. Fully interactive. Fast (parallel code). Extensive automation. Standalone windows executable with installer. Online help. 100% LabVIEW. Well received by users.
- Disadvantages: Windows only (also runs on virtual PC).
- DeerAnalysis (Gunnar Jeschke, EPR@ETH software, ETH)
- The gold standard of course is DeerAnalysis By Gunnar Jeschke.
- Advantages: Rich heritage and more features for unusual cases. Excellent manual.
- Disadvantages: Requires a MATLAB license.
- DEERLab. A modern Matlab toolbox for Dipolar EPR spectroscopy. (On Github)
- Open source collection of the most popular dipolar analysis methods. MIT license.
- Currently in pre-release, but will have great potential once polished.
- Advantages: Rich collection of functions. Well tested and documented code (asymptotically).
- Now migrated to Python! (originally in Matlab)
- DEfit (Ilker Sen, Jean Chamoun, P. Fajer, alternative download, Florida State)
- DEFit is a Double Electron-Electron Resonance (DEER) data analysis program which assumes models of overlapping Gaussian shaped distance distributions between spins. DEFit utilizes Monte Carlo/SIMPLEX curve fitting.
- Advantages: Not tested.
- Disadvantages: Requires MATLAB license. Limited documentation. Last updated in 2009.
- PD_Tikhonov (ACERT software, Cornell)
- Includes all necessary functions required to extract the pair distributions from pulsed ESR experiments using the Tikhonov regularization method.
- Advantages: Not tested.
- Disadvantages: Requires MATLAB license and experience, requires regularization toolkit, limited documentation. Last updated ~2005.
- DPD-pkg (ACERT software, Cornell)
- An upgraded version of PD_Tikhonov package since both TIKR and MEM are included.
- Advantages: Not tested.
- Disadvantages: Requires MATLAB license and experience, requires regularization toolkit, limited documentation. Last updated ~2008.
- GLADD/DD (Eric Hustedt, Vanderbilt)
- DD is a GUI version of the GLADD program (Matlab commandline) for fitting DEER data.
- Advantages: Not tested.
- Disadvantages: Requires MATLAB license. Limited documentation.
- DeerA2012 (R Stein, Vanderbilt)
- Quote from the description: "The program will fit one or more DEER echo decays to a user-defined sum of Gaussian distributions. The background correction is part of the fit routine. The program will pick the statistically significant fit, though the best results from each number of Gaussian distributions are also saved. The results and fits are saved in an Excel File (currently uses ActiveX for saving results).
- Advantages: Not tested.
- Disadvantages: Requires MATLAB license. Limited documentation.
Q: What is expected from users of the program?
A: The LongDistances software is ready to be used by anyone, but there are some missing features (e.g. the Data Manager) that will be implemented at a later point. There could also be slight changes in features and algorithm in the future. The code is stable and works exceptionally well for typical data and has been in use in several laboratories for a long time.
Users are strongly encouraged to provide feedback on all aspects of the program, such as:
- Report bugs, crashes, and other unexpected behavior
- Suggest features and changes
- Point out problems with usability
- Point out problems with pathological data
- Point out problems with incorrect definitions of parameters
- etc.
Q: What are the various algorithms?
A: All algorithms were written from scratch and implemented in LabVIEW. During development, many dozens of variations were extensively tested, tuned, and compared. The current code is highly optimized for accurate results under a wide variety of scenarios with real, as well as simulated data sets. At the same time, everything is optimized for speed and can take advantage of all available CPU cores with highly parallel code architecture.
Algorithms were developed for many aspects of the program, such as
- Automatic phase correction
- Near instantaneous background and result estimation (what you see before optimizing the background or fitting)
- Tikhonov (optionally non-negative) regularization in milliseconds
- Model-free fitting
- Model-based fitting
- Co-fitting of background (even for model-free!)
- Very high quality full kernel generation in sub-milliseconds.
- Extensive error analysis
- Fully interactive and reactive user interface.
Most of the algorithms are described in detail here.
LabVIEW™ Is a trademark of National Instruments Corporation. MATLAB is a trademark of The MathWorks, Inc