Tips & Tricks for using Melanie-MS

1-  Generating data for display and analysis by MSIGHT

1.1 LC-MS data

Higher quality data for profiling of a complex mixture are obtained by acquiring LC-MS only data first (i.e. no data-dependent MS/MS scans). This ensures that MS spectra are taken at regular time intervals and allows to perform proper integration of the elution profiles, an essential prerequisite for quantitative measurements. On the other hand, it is quite redundant to acquire data at rates that are too fast. This generates large numbers of spectra, hence huge image files, while the quality of the data is not massively improved.

We have found that the following settings yield good data :

QSTAR (QTOF) :

·  Setup : nano-LC at 0.2 ul/min ; typical run time 45-60 min ( gradient length 20-35 min to go from 1% to 30% MeCN)

·  scan m/z 400-1200 in 3.0 seconds

·  These chromatographic conditions results peptide peaks of about 25-40 sec. Accumulating one full scan spectrum every 3 s yields higher quality spectra than usual 1.0 s scans, hence higher sensitivity and higher quality images.

·  The sampling of about 10 points / LC peak is reasonable. Also 60 min gradients can be used.

ION TRAPS :

·  scan m/z 400-1200

·  Add several spectra to obtain full scan time around 2-3 s

·  Similar mass ranges and other considerations like for the QSTAR apply.

·  However, gain control on traps causes scan times to vary considerably depending on signal intensity, a major problem for quantitative measurements. More scans are acquired when there are intense signals. If accurate quantitation is required, efforts should be made to make scans as regularly as possible on the time scale.

1.2 LC-MS + data dependent MS/MS

Of course it is great to be able to get some proteins identified anyway, while doing the profiling. This is possible, at the expenses of some data quality. We have used the following settings :

QSTAR (QTOF) :

·  Setup : nano-LC at 0.2 ul/min ; typical run time 45-60 min ( gradient length 20-35 min to go from 1% to 30% MeCN)

·  Survey scan : m/z 400-1200 in 3.0 seconds

·  MS/MS on the most intense peak accumulated for 4.0 s

·  Threshold for choice of parent ion : very low (i.e. 3 cps, normally around 15)

Survey scans are measured in conditions similar to what described above. Only one parent at the time is selected for analysis, to limit the interval between full scans. Setting a low threshold causes the instrument to pick precursors all the time, and this in turn results in a more homogenous density of survey scans in time (and, of course, a lot of garbage CID spectra).

ION TRAPS

·  scan m/z 400-1200

·  Add several spectra to obtain full scan time around 2-3 s

·  CID on most intense precursor ; limit accumulation to max 3-4 secs

·  Threshold for CID : as low as possible

.

Similar considerations as for the QSTAR instruments apply.

2- Exporting data to text format (.txt)

·  Open a file in Analyst.

If the data are simple MS

·  Launch the ExportToText…dll script.

·  Choose a threshold slightly above level of noise in your data (! Noise levels are higher in files acquired with longer scan times).

·  Do not centroid data, unless you want. Image size is the same for centroided and noncentroided data, and by centroiding some signals could be lost.

If the data are MS-MS/MS

·  double click on the green arrow below the m/z axis of the chromatogram. This displays the TICs for all scans (“experiments”) separately.

·  Activate the pane of the survey scan (clicking inside).

·  Launch the ExportToText…dll script.

·  Chose an appropriate threshold (20% above average noise level). Do not centroid.

2-  Importing data into MSIGHT

·  Open MSIGHT.

·  Go to workspace, open a project>MS-RUN>Add>Imported MS-run ( in general : follow MSIGHT documentation )

·  Set the step size. This is the mass step for sampling the data in the m/z dimension. If you want to preserve the whole resolution of your data, this should not be set too large. For the QSTAR : under the settings described usually choose 0.05 (mass step is 0.01 in a normal scan ) .

·  For lower resolution data step size can be larger (0.1-0.2).

·  Minimum and maximum are mass limits in case you want to visualize only part of the mass range ( we find usually that a mass range 400-1200 is sufficient for ESI data)

! IMAGE size is a function of number of scans and mass range ! For example at one scan every 3 s, a 45 min method will yield 900 scans. On a QSTAR with a mass step size of 0.06 Da, and a scan range of 800 (400-800 a.m.u. ), there are 12 millions data points…

Every one of these parameters should be judiciously chosen to avoid generating useless data and huge, non-manageable image files.

(Daniel please correct and complete this !)