ADAPT-NMR (Assignment-directed Data collection Algorithm utilizing a Probabilistic Toolkit in NMR) is a fully automated method for integration of protein NMR data collection, chemical shift assignment, and secondary structure determination.

BACKGROUND

ADAPT-NMR optimizes a fast data collection path involving multiple reduced dimensionality experiments so as to most efficiently realize complete resonance assignment and secondary structure determination of a protein while spectra are being collected on-the-fly.

With the sample in the NMR probe, ADAPT-NMR directs the NMR spectrometer to collect the optimal 2D plane of the most favorable experiment and then analyzes the resulting spectra to determine the positions of the peaks and to update the spins systems, chemical shift assignment, and the secondary structure of the protein. The procedure continues until the collected data support comprehensive peak identification, chemical shift assignments at the desired level of completeness, and protein secondary structure determination.

The process does not involve manual intervention and in favorable cases requires less than a day.

We also offer ADAPT-NMR Enhancer for visualization and verification of ADAPT-NMR results, and provides for iterative runs of ADAPT-NMR with user-edited peaklists.

DOWNLOAD

ADAPT-NMR has been developed for both Agilent(Varian) and Bruker NMR spectrometers runnning Linux.  It is freely available for download (see below).

Varian (Agilent)

To install ADAPT_NMR follow these steps:

1) Download and save the ADAPT-NMR package for Agilent(Varian) file onto your Linux computer. (e.g. in your home directory: /home/marco)

2) Download in install NMRPipe.

3) cd to your vnmrsys directory. (e.g. cd /home/marco/vnmrsys)

4) Run the untar command from within this directory. (e.g. tar xvf /home/marco/adapt.tar)

5) Go to psglib directory and compile the pulse programs. (e.g. cd /home/marco/vnmrsys/psglib ; seqgen NEW*)

6) Go to adapt_nmr/libraries directory and install the MATLAB libraries - this will take a few minutes.

    (e.g. cd /home/marco/vnmrsys/adapt_nmr/libraries ; and then follow the instruction in install.txt).

7) Go to input_files directory (e.g. cd /home/marco/vnmrsys/adapt_nmr/input_files) and update the first two lines of parameters.txt file as follows:

-adapt_home : should be the directory where adapt_nmr was installed, e.g.: /home/marco/vnmrsys/adatpt_nmr

-nmrpipe_source : should point to the required NMRPipe source file

8) Follow instructions in strategy.README in the input files directory to run ADAPT_NMR. Also, a full description of input files and output files has been provided in the website.

 

Bruker

To install ADAPT_NMR follow these steps:

1) Download and save the ADAPT-NMR package for Bruker file onto your computer (e.g. on your home directory: /home/whlee)

2)Download and install NMRPipe.

3) Create and cd to your adapt_src directory (e.g. mkdir /home/whlee/adapt_src; cd /home/whlee/adapt_src)

4) Run the unzip command from within this directory (e.g. unzip /home/whlee/adapt_B.zip)

5) Run install.py (e.g. python install.py)

6) Go to adapt_src/adapt_nmr/libraries directory and install the MATLAB libraries - this will take a few minutes

(e.g. cd /home/whlee/adapt_src/adapt_nmr/libraries ; and then Follow the instruction in readme.txt)

7) Go to essentials directory (e.g. cd /home/whlee/adapt_src/essentials) and update the first two lines of parameters.txt file as follows:

-adapt_home : should be the directory where adapt_nmr was installed, e.g.: /home/whlee/adapt_src/adapt_nmr

-nmrpipe_source : should point to the required to NMRPipe source file.

8) Follow instructions in strategy_B.README in the essentials directory to run ADAPT_NMR.

 

MANUAL

ADAPT-NMR is a comprehensive software platform that integrates many other computational methods developed at NMRFAM such as HIFI and PINE.  Below is described in detail the setup, inputs and outputs of the platform.  If you have any further questions, feel free to contact the NMRFAM Director.

Setup

To setup an ADAPT-NMR run on a compatible spectrometer requires several text file inputs to be edited, as described below.

 

Setup of ADAPT-NMR on Varian (Agilent) Spectrometers

Create a directory for running your experiments (e.g. <ADAPT_NMR_on_my_protein>).

Copy the parameters.txt, nmrpipe.par, ORTHO_list.txt, ADAPT_list.txt files from the ~/vnmrsys/adapt_nmr/input_files directory to the <ADAPT_NMR_on_my_protein> directory you just created (also copy a text file with the sequence for your protein).

Update the parameters.txt file accordingly (see Input Files/Parameter File).

Calibrate pw/tof for your protein sample and save the values in the probefile.

Running ADAPT_ORTHO to collect and process orthogonal planes

In vnmrJ, run quick 2D experiments to establish spectral window and offset in the indirect dimension for:

N15 (run N15-HSQC, use sw1, dof2 for N15-HSQC)
CO (run HNCO, use sw1, dof for HNCO HNCACO)
CA (run HNCA, use sw1, dof for HNCA HNCOCA)
CB (run CBCACONH, use sw1, dof for CBCACONH, HNCB)

NOTE: do NOT make the spectral window in the indirect dimension too tight or the peaks may fold in the tilted planes

Edit ORTHO_list.txt file: Enter the sw1, dof/dof2 values calibrated above (replace values given in the example input file).

Edit nmrpipe.par file: From quick HNCO above, set parameters for extracting the amide region in the direct dimension for nmrpipe processing (remember that the values are in points and the final total size for the direct dimension is 2048). Enter these into the nmrpipe.par file.

Apart from the N15-HSQC, for all other experiments the phasing parameters for the direct proton dimension should be identical, so you can just enter the rp value (lp should be zero) you get from the quick 2D HNCO you ran above into the nmrpipe.par file

Go to vnmrj, cd into the <ADAPT_NMR_on_my_protein> directory and type "ADAPT_ORTHO".

All orthogonal planes listed in ORTHO_list.txt will be automatically collected and processed (you can look at spectra in vnmrj as they are being collected, but don't change experiment or quit vnmrj or you'll interrupt the run).

If you need to adjust parameters and reprocess some of the spectra, edit the ft_ortho.com file that is saved with each experiment and run the "run_nmrpipe" macro to reprocess each spectrum. NOTE: The processing parameters from the 0 degree orthogonal plane will be used to process the tilted planes.

Once all orthogonal planes are collected and processed properly, you are ready to run ADAPT_NMR

Running ADAPT_NMR

Edit ADAPT_list.txt file (very similar to ORTHO_list.txt)

Type ADAPT_NMR in vnmrj command line and let the spectrometer and computer do their work !!

At the end, the results should be located inside the output_files directory.


Setup ADAPT-NMR on Bruker spectrometers

Create a directory for running your experiments (e.g. <ADAPT_NMR_on_my_protein>).

Copy the parameters.txt, nmrpipe.par, ORTHO_list.txt, ADAPT_list.txt files from the ~/adapt_src/essentials directory to the <ADAPT_NMR_on_my_protein> directory you just created (also copy a text file with the sequence for your protein).

Update the parameters.txt file accordingly (see Input Files/Parameter File)

Running ADAPT_ORTHO to collect and process orthogonal planes

In terminal, move your current directory into the <ADAPT_NMR_on_my_protein>.
$ cd <ADAPT_NMR_on_my_protein>.

Run TopSpin, and run quick 2D experiments to establish spectral window and offset in the indirect dimension for:

N15 (run N15-HSQC with modified pulse program 'hsqcfpf3gpphwg_hifi.khu', use SW[ppm], O3P[ppm] for N15-HSQC)
CO (run HNCO with modified pulse program 'hncogpwg3dsc_hifi.khu', use SW[ppm], O2P[ppm] for HNCO HNCACO)
CA (run HNCA with modified pulse program 'hncagpwg3dsc_hifi.khu', use SW[ppm], O2P[ppm] for HNCA HNCOCA)
CB (run CBCACONH with modified pulse program 'cbcaconhgpwg3dsc_hifi.khu', use SW[ppm], O2P[ppm] for CBCACONH, HNCB)

NOTE: do NOT make the spectral window in the indirect dimension too tight or the peaks may fold in the tilted planes

Edit ORTHO_list.txt file: Enter the SW, O2/O3 values calibrated above (replace values given in the example input file).

Edit nmrpipe.par file: From quick HNCO above, set parameters for extracting the amide region in the direct dimension for nmrpipe processing (remember that the values are in points and the final total size for the direct
dimension is 2048). Enter these into the nmrpipe.par file.

Apart from the N15-HSQC, for all other experiments the phasing parameters for the direct proton dimension should be identical, so you can just enter the rp value (lp should be zero) you get from the quick 2D HNCO your ran above into the nmrpipe.par file

Go to topspin after cd into the <ADAPT_NMR_on_my_protein> directory, type "ADAPT_ORTHO_run".

All orthogonal planes listed in ORTHO_list.txt will be automatically collected and processed (you can look at spectra in topspin as they are being collected, but don't change experiment or quit topspin or you'll interrupt the run).

If you need to adjust parameters and reprocess some of the spectra, edit the ft_ortho.com file that is saved with each experiment and run the "run_nmrpipe" macro to reprocess each spectrum. NOTE. The processing parameters from the 0 degree orthogonal plane will be used to process the tilted planes.

Once all orthogonal planes are collected and processed properly, you are ready to run ADAPT_NMR.

Running ADAPT_NMR

Edit ADAPT_list.txt file (very similar to ORTHO_list.txt)

Type "ADAPT_NMR_run" in topspin command line and let the spectrometer and computer do their work !!

At the end, the results should be located inside the output_files directory.

Inputs

For the setup of ADAPT-NMR described above, several local files must be established as input by the user:
  • A parameter file
  • A sequence file
  • Optional sidechain peaklists
  • Internal files

These files are described in more detail below.

 

Parameter file

The parameter file, parameters.txt, is a text file that should be located at the directory that you are running ADAPT-NMR. It includes all the user information that ADAPT-NMR needs to run.  See Example <need link here>.

Most of the parameters have been optimized for a typical protein and do not need to be modified in most cases. However, there are required parameters (n_residue, sequence_file, and sequence_letter) that need to be set by the user at each run.

The following table describes all the parameters in the file and indicates if any modification is required or suggested. We suggest no modification to the values flagged as “Optimized” unless if you are unhappy with the final result. You can always change those parameters after a full run of ADAPT-NMR and the result will be updated accordingly.

Parameter

Default Value

Acceptable Values

Modification Suggestion

Description

n_residue

N/A

Any

required

Number of amino acids in the protein

sequence_file

N/A

Relative path of the sequence file

required

Describes location of amino acid sequence file

sequence_letter

N/A

3 or 1

required

Specifies if the sequence file you provided is in 1- or 3-letter format

hcconh_peak_list_file

0

Relative address of the peak list file

Optional

3D HBHA(CO)NH peaklist for sidechain assignments

cconh_peak_list_file

0

Relative address of the peak list file

Optional

3D H(CCO)NH peaklist for sidechain assignments

hcch_tocsy_peak_list_file

0

Relative address of the peak list file

Optional

3D HCCH-TOCSY peaklist for sidechain assignments

hbhaconh_peak_list_file

0

Relative address of the peak list file

Optional

3D HBHA(CO)NH peaklist for sidechain assignments

assignment_level

0.95

0-1

Optimized

 

Data collection stop criteria. The default value is set to reach at least 95% of assignment completeness.

window_size_peak_picking

2

1-5

Optimized

Peak picking algorithm parameter. Setting the default value as two, it recognize a data point as a Local Maximum if its height is higher than two adjacent data point in any direction.

min_peak_distance

3

1-5

Optimized

Peak picking algorithm parameter.  If a peak is less than ‘min_peak_distance’ data point away from a larger peak it should satisfy other requirement to be recognized as a peak. It has been optimized to minimize the noise artifact.

max_1d_aligning_shift

3

1-10

Optimized

Resonance alignment algorithm parameter that determines the maximum allowed shift across tilted planes.

max_hsqc_peaks

1.3

Any

 

Optimized

HSQC peak picking parameter, setting the upper limits for the ratio of the number of peaks found in the 15N-HSQC, divided by the number of amino acids.

min_hsqc_peaks

1.1

Any

 

Optimized

HSQC peak picking parameter, setting the lower limits for the ratio of the number of peaks divided by the number of amino acids.

xsens_disp

3

1-10

Optimized

A parameter in the dispersion function calculation that defines the ideal dispersion. Two peaks are considered dispersed in the direct dimension if they are more than ‘xsens_disp’ data point away in the direct dimension.  

ysens_disp

3

1-10

Optimized

Two peaks are considered dispersed in the indirect dimension if they are more than ‘ysens_disp’ data point away in the direct dimension.  

water_centered_flg

0

0 or 1

Optional

 

 

Protein sequence input file

The protein sequence input text file should contain the amino acid sequence in a single column, using either 1- or 3-letter amino acid codes.

Examples:

M   
T   
E   
V   
Y   
D   
L   
E   
I   
T   
T   
N   
A   
.......
.......
or
MET   
THR   
GLU   
VAL   
TYR   
ASP   
LEU   
GLU   
ILE   
THR   
THR   
ASN   
ALA
.......
....... 

Optional sidechain peaklists

Currently, ADAPT-NMR supports sidechain assignment if the user provides a path to sidechain peaklists in the parameter file (see above). The peak list should be provided in the "UNASSIGNED" Sparky format (see below).

The following sidechain experiments are supported by ADAPT-NMR:

     HBHA(CO)NH
     HCCH-TOCSY
     H(CCO)NH
     C(CO)NH
     13C-HSQC

We are planning to include adaptive collection and assignment of sidechain data in the next release of ADAPT-NMR.

 

Internal files

The internal files described below are generated or updated by ADAPT-NMR at runtime and need not be altered (modification of these files is NOT recommended):

ADAPT_list.txt: List of experiments to be collected.  Default list is "HNCO,HNCA,HN(CO)CA,HN(CA)CO,HNCB,CBCA(CO)NH,C(CO)NH,HBHA(CO)NH,HA(CO)NH,H(CCO)NH,C13-HSQC".

input_hifi.txt: Collected experiments and 2D planes. This file contains the information about the collected spectra (experiment type and the tilted planes collected) and the location of their processed data (ft2 files). It is essential for communication of data collection module and data analysis module.

suggested_plane.txt: Suggested experiment and tilted plane. This file, generated by the data analysis module, suggests the experiment and the tilted plane to be collected, to the data collection module

ftHF.com: NMRPIPE script for processing tilted planes.

ft0.com: NMRPIPE script for processing orthogonal planes

 

NMRSTAR chemical shift assignment format

protein_nmrstar21.str: Chemical shift assignments in NMRSTAR (BMRB) 2.1 format.  See example below, or visit BMRB for more details.

NMRSTAR 2.1 example

protein_nmrstar31.str: Chemical shift assignments in NMRSTAR (BMRB) 3.1 format. See example below, or visit BMRB for more details.


Output Files

Chemical shift assignments

protein_backbone_assignment.txt: a probabilistic assignment of amide chemical shift for every residue. This means that it is possible for some residues to have more than one candidate spin systems. In these cases, an associated probability is assigned to each candidate spin system assigned to the residue. The sum of all probabilities always adds to one. See Examples <need link>.

sidechain_table.txt: if the user provides sidechain peak lists (see Input Files/Parameter file), ADAPT-NMR produces a probabilistic assignment for every atom in each residue. This means that it is possible for some atoms to have more than one candidate chemical shift assignment. In these cases, an associated probability is assigned to each candidate assigned to an atom. The sum of all probabilities always adds to one. See Examples <need link>.

<experiment name>.list: spectra peaklists in SPARKY peaklist format.  See Examples <need link>.

protein_nmrstar21.str: Chemical shift assignments in NMRSTAR (BMRB) 2.1 format. See Examples <need link>, or visit BMRB.

protein_nmrstar31.str: Chemical shift assignments in NMRSTAR (BMRB) 3.1 format. See Examples <need link>, or visit BMRB.

protein.jpg: A jpg figure demonstrating assignment probability for each residue. See Examples <need link>.

pecan_fig.jpg: A jpg figure demonstrating the likely secondary structure of each residue. See Examples <need link>.

Internal files

input_hifi.txt: Collected experiments and 2D planes.  This file contains the information about the collected spectra (experiment type and the tilted planes collected) and the location of their processed data (ft2 files). It is absolutely important for communication of data collection module and data analysis module.

suggested_plane.txt: Suggested experiment and tilted plane. This file, generated by the data analysis module, suggests the experiment and the tilted plane to be collected, to the data collection module.

ftHF.com: NMRPIPE script for processing tilted planes.

ft0.com: NMRPIPE script for processing orthogonal planes.

EXAMPLES

Protein sequence input file

The protein sequence input text file should contain the amino acid sequence in a single column, using either 1- or 3-letter amino acid codes.

Examples:

M   
T   
E   
V   
Y   
D   
L   
E   
I   
T   
T   
N   
A   
.......
.......
or
MET   
THR   
GLU   
VAL   
TYR   
ASP   
LEU   
GLU   
ILE   
THR   
THR   
ASN   
ALA
.......
....... 

CITATION

Bahrami A, Tonelli M, Sahu SC, Singarapu KK, Eghbalnia HR, Markley JL. Robust, integrated computational control of NMR experiments to achieve optimal assignment by ADAPT-NMR. PLoS One. 2012;7(3):e33173. doi: 10.1371/journal.pone.0033173. Epub 2012 Mar 12. PubMed PMID: 22427982; PubMed Central PMCID: PMC3299752.

Lee W, Hu K, Tonelli M, Bahrami A, Neuhardt E, Glass KC, Markley JL. (2013) Fast automated protein NMR data collection and assignment by ADAPT-NMR on Bruker spectrometers. Journal of Magnetic Resonance. Aug 30;236C:83-88. doi: 10.1016/j.jmr.2013.08.010 (JMR link).08.010 (JMR link)

Lee W, Bahrami A, Markley JL. ADAPT-NMR Enhancer: complete package for reduced dimensionality in protein NMR spectroscopy. Bioinformatics. 2013 Feb 15;29(4):515-7. doi: 10.1093/bioinformatics/bts692. Epub 2012 Dec 7. PubMed PMID: 23220573; PubMed Central PMCID: PMC3570218.

Dashti H, Tonelli M, Markley JL. ADAPT-NMR 3.0: utilization of BEST-type triple-resonance NMR experiments to accelerate the process of data collection and assignment. J Biomol NMR. 2015 Jul;62(3):247-52. doi: 10.1007/s10858-015-9950-7. PubMed PMID:: 26021595; PubMed Central PMCID: PMC4687732.