Agilent Technologies Spectrum Mill MS G2721AA User manual

Agilent Technologies
Agilent G2721AA Spectrum Mill
MS Proteomics Workbench
Quick Start Guide
A guide to the Spectrum Mill workbench
Use this reference for your first steps with the Spectrum Mill workbench.
What is the Spectrum Mill MS Proteomics Workbench?
The Spectrum Mill software is a collection of tools for high-
throughput processing of MS and MS/MS spectra to provide protein
and peptide identifications and relative quantitation. The
Spectrum Mill workbench can do in minutes what could take you
hours or days to do manually.
This software includes bioinformatics tools for:
•Spectral preprocessing for MS/MS and MS-only data, including
spectral quality filtering and precursor charge state assignment
for MS/MS spectra
•Protein database search, including searches for both MS/MS and
MS-only spectra
•Data review, validation, and comparison for large data sets,
including those from two-dimensional LC/MS/MS analyses and
differential expression studies
•De novo sequencing
•Identification of mutations, post-translational modifications,
and chemical modifications
•Differential expression quantitation
The Spectrum Mill workbench accepts Agilent data, as well as data
from other vendors’ mass spectrometers. The software executes on
a web server, and you access the program via a web browser
window.

2 Spectrum Mill Workbench Quick Start Guide
New features in Spectrum Mill workbench version A.03.03
The following are key changes to the user interface and functionality:
Support for more
instruments and
data types
Now supported:
•Agilent Q-TOF
•Electron transfer dissociation (ETD) data from Agilent ion trap
•MALDI-TOF-TOF data (via *.mgf file, requires license for Generic Data
Extractor)
•Thermo Fisher Scientific LTQ, LTQ FT, and LTQ Orbitrap, in various
dissociation modes (requires a license for Thermo Data Extractor, as well as
co-resident Thermo Fisher Scientific Xcalibur software that is equivalent to or
newer than the version you used to acquire the data)
Data Extractor Ability to disable similarity merging, to improve coverage and to detect more
low-level peptides for simple samples (for example, a single protein)
MS/MS Search Capability to search peptides that have a charge up to +7
New Maximum ambiguous precursor charge, which allows you to choose the
maximum charge state that you want MS/MS Search to consider when the Data
Extractor could not discern the charge state for the peptide
Capability to use protein pI to filter candidate proteins from the database
Improved scoring schemes, including Dynamic peak thresholding, which is a
scoring enhancement that enables identification of more low-abundance and
short-chain peptides
Explicit “mass gap” search that allows you to search for an unexpected or
unknown modification
Ability to automatically create a file of previously validated hits from the MS/MS
Search page, without using the Tool Belt page
MS/MS Autoval-
idation
Addition of +5 charge state to the autovalidation rules
Capability to perform autovalidation across data files in multiple subdirectories,
with all peptides contributing to the protein score
Addition of peptide pI as a validation criterion for samples where the pI is known
(for example, from offgel fractionation)

Spectrum Mill Workbench Quick Start Guide 3
Protein/Peptide
Summary
Spectrum Viewer support for c- and z-ions, which are prominent in ETD spectra
Retention time in peptide summaries
Fragmentation mode in peptide summaries, so you can ascertain the mode for
each spectrum (for example, CID versus ETD)
Enhanced protein grouping based upon shared peptides
Ability to align sequences of proteins within a protein group
Ability to sort and filter by peptide pI
New capability in the Spectrum Viewer to type your own sequence and have it add
to the list of available sequences that the software shows when you click the Rank
arrow buttons
Ability to sort peptides by the numerical position of their first amino acid in the
protein sequence
Capability to distinguish in results tables between true mass errors and mass
differences produced by peptide modifications
Support for more
amino acid modi-
fications
Support for Deamidated (N) and Deamidated (Q) as variable modifications. The
software displays the first by default; system administrators can add the second by
copying a definition from smconfig.misc.xml to smconfig.custom.xml.
Expanded iTRAQ support, to include data from the Generic Extractor and the
*.raw (LCQ/LTQ) Extractor. The latter uses merged MS3spectra.
Support for both original iTRAQ (which assumes complete labeling) and iTRAQ
Partial-mix, which allows you to check for incomplete labeling
Miscellaneous New Easy MS/MS Search form that simplifies use of the software and streamlines
the data processing steps
New Multiple Sequence Aligner utility to align sequences of proteins from a
database and highlight amino acid differences
Support for Web-based interaction with GeneSpring MS
Easier, faster input of mass lists into Manual PMF, with MH+/M selection that
allows you to input either protonated or neutral species (directly compatible with
output from a variety of instrument data systems)
Faster selection of data folders
Sherenga de novo sequencing support for the Agilent Q-TOF and ETD data from
the Agilent ion trap

4 Spectrum Mill Workbench Quick Start Guide
The ability to store iTRAQ correction factors for multiple iTRAQ batches and to
apply them to the iTRAQ calculations for specific data sets
Expanded capabilities for Peptide Selector
Updated instrument selection lists
The default Spectrum Mill font size is now “smaller”, so you may use the default
“medium” size in your browser and the Spectrum Mill software still uses “smaller.”

Spectrum Mill Workbench Quick Start Guide 5
Spectrum Mill manuals and online help
Scientists Familiarization Guide Follow step-by-step instructions to process example data.
Application Guide Learn details to use the software.
Online Help Consult the online help for in-depth information not given in the
Application Guide. To access help, click links on the home page, Help buttons on
Spectrum Mill forms, or blue dividing bars on Spectrum Mill forms.
Quick Reference Card After you are familiar with the software, consult this card
for an overview of the steps to process MS/MS data.
Thermo Ion Trap Data Extractor Quick Start Guide Get a quick overview of the
optional data extractor for Thermo Fisher Scientific ion trap data files.
Generic Data Extractor Quick Start Guide Get a quick overview of the optional
Generic Data Extractor for peak list files.
QSTAR Data Extractor Quick Start Guide Get a quick overview of the optional
Applied Biosystems/MDS Sciex QSTAR Data Extractor.
System
Administrators
Installation Guide Use this guide to install the software on the server.
Application Guide See the following chapters:
•Chapter9:System Administration
Get an overview to install databases and perform other system administration tasks.
• Chapter 10: Files Created during Spectrum Mill Data Processing
Refer to this chapter to troubleshoot data processing, to selectively remove
parts of the processing, or to decide which files to archive.
Online Help From any Help page, click links under For System Administrators:
•ProteinDatabases(link to millhtml\SM_instruct\faman.htm)
Learn details to install databases, create indices, and create subset databases.
• Server Administration (link to millhtml\SM_instruct\servadmn.htm)
Learn details to perform other system administration tasks.

6 Spectrum Mill Workbench Quick Start Guide
Setting up the Spectrum Mill server
See the Spectrum Mill MS Proteomics Workbench Installation Guide. If you wish
to update from a previous version of the Spectrum Mill workbench, see Chapter 2.
Note that the server name cannot have an underscore.
Setting up your client PC
Operating system Check that you have one of the following:
•Windows®2000 Server, SP4
•Windows Server 2003, SP1
•Windows XP, SP2
1Select Start >Run....
2In the Open box, type winver.
Browser Check that your web browser is Internet Explorer 6.0 (IE 6.0) or Internet Explorer
7.0 (IE 7.0).
1Open Internet Explorer.
2Select Help > About Internet Explorer.
Java Check that Internet Explorer on the client is configured for Sun JRE 1.5.x (also
called J2SE 5.0) or Java SE 6. You can get a compatible version of JRE from your
Spectrum Mill installation CD.
1With IE 6.0 or IE 7.0, select Tools > Internet Options....
2Click the Advanced tab.
3Scroll down to the Java (Sun) section to view the version.
Screen resolution If supported on your monitor, use the Control Panel to set screen resolution to
1280 x 1024, with True Color (32 bit).
1In Control Panel, double-click Display.
2Click the Settings tab.
3Adjust the settings if necessary.
NOTE At the time of release, Windows Vista is not yet supported. If you wish to use Vista, please
contact Agilent to check the support status of Vista.

Spectrum Mill Workbench Quick Start Guide 7
Stored pages Disable caching.
1With IE 6.0 or IE 7.0, select Tools > Internet Options....
2Click the General tab.
3Do one of the following:
•If you have IE 6.0, under Temporary internet files, click the Settings...
button. Then click Every visit to the page.
•If you have IE 7.0, under Browsing history, click the Settings... button. Then
click Every time I visit the webpage.
Cookies Enable cookies so you can save form settings.
1With IE 6.0 or IE 7.0, select Tools > Internet Options....
2Click the Privacy tab and set to Low, or click the Advanced button and enable
cookies.
3For more detailed instructions to enable cookies, search Internet Explorer help
for cookies.
4If the drop-down menus in the Spectrum Mill workbench appear empty, you
may need to add the Spectrum Mill server as a trusted site. Search Internet
Explorer help for trusted site.
Active scripting Enable JavaScript (Active Scripting).
1With IE 6.0 or IE 7.0, select Tools > Internet Options....
2Click the Security tab.
3Select Local Intranet.
4Click the Custom Level... button.
5In the Security Settings dialog, scroll down until you see the Scripting section.
6Under Active Scripting, click the Enable option.
Browser text size Especially for lower-resolution monitors (e.g., 1024 x 780), set a smaller text size to
avoid excessive scrolling.
1Do one of the following:
•With IE 6.0, select View > Text Size.
•With IE 7.0, select Page > Text Size.
2Set to Smaller.

8 Spectrum Mill Workbench Quick Start Guide
Running the Spectrum Mill workbench from the client
Open Internet Explorer and point to the URL for the Spectrum Mill server. This
has the general format server/millhome.htm. If you are not sure of the URL, ask your
system administrator.
Setting up Agilent TOF analyses for use with the Spectrum Mill
workbench
When you set up data acquisition parameters on the Agilent TOF, it is critical that
you establish analysis settings that are compatible with Spectrum Mill processing.
For details, see the chapter on processing MS-only data in the Spectrum Mill MS
Proteomics Workbench Application Guide.
Software Status Bulletin and software patches
A list of known problems for the Spectrum Mill workbench, along with possible
solutions, is found on the Agilent Web site.
1Go to http://www.agilent.com/chem/.
2On the left, click Technical Support.
3Under Downloads and Utilities, click Status Bulletins and Patches.
4On the Software Status Bulletins and Patches page, click Spectrum Mill.
5If you are prompted to log in, and you are registered on the Agilent Web site,
type your login name and password, then click Enter. If you are not registered,
click the link to register.
6On the Software Status Bulletins and Patches page, click Spectrum Mill.
7When prompted for your Spectrum Mill password, type specmill (all lower
case) into the text box, then click SUBMIT.
8View the Software Status Bulletin, or download a software update.

Spectrum Mill Workbench Quick Start Guide 9
Spectrum Mill modules
Tool type Module Function ApplicationGuide
Chapter(s)
Spectral
preprocessing
Data Extractor Extract spectra from data files. Processing depends
on data type and may include averaging,
centroiding, filtering by quality, assigning precursor
charge, and calculating spectral features.
1, 6
Interpretation MS/MS Search Search MS/MS spectra against databases. 1
Spectrum Matcher Match MS/MS spectra against other MS/MS
spectra.
1
de novo Sequencing Use Sherenga de novo sequencing to interpret
MS/MS spectra.
1, 3
PMF Search Search MS-only spectra against databases. 6
Result summary Protein/Peptide Summary Summarize, review, filter, and validate results from
MS/MS searches.
1, 2
Spectrum Summary Sort and classify MS/MS spectra based on spectral
characteristics.
1
de novo Summary Summarize and review de novo interpretations. 1, 3
PMF Summary Summarize and review results from MS-only
searches.
6

10 Spectrum Mill Workbench Quick Start Guide
Utilities Tool Belt Use tools to speed processing, examine parameters
and search statistics, stop processes, etc.
7
Protein Databases Create indices for databases so searches run faster,
create subset databases, etc.
9
Build TIC Visualize how much of your sample has been
interpreted. Locate spectra that indicate potential
peptide phosphorylation sites.
1
Peptide Selector Perform theoretical digestion and then select
peptides that meet certain criteria (e.g., likely to
produce high-quality MS/MS spectra).
8
Multiple Sequence Aligner Align amino acid sequences of proteins from a
database and highlight the amino acids that differ
8
MS Digest Perform theoretical enzymatic digestion and
calculate masses of peptides that result.
8
MS Edman Search text fields in protein databases, with or
without peptide mass filtering.
8
MS Product Calculate theoretical fragment ion masses from
peptides.
8
MS Comp Calculate amino acid composition for a peptide,
given peptide mass and immonium ions.
8
MS Isotope Calculate and view isotope patterns of peptides. 8
Tool type Module Function ApplicationGuide
Chapter(s)

Spectrum Mill Workbench Quick Start Guide 11
Roadmap for MS/MS data processing
See Chapter 1 in the Application Guide for details.
f
Preprocess spectral files(Data
Extractor page)
Set search parameters and run
searches (MS/MS Search page)
Validate highest-quality results
automatically (Autovalidation
page)
Validate medium-quality results
manually (Protein/Peptide
Summary page)
Summarize valid results
(Protein/Peptide Summary
page)
Do you want to
process the data
further?
Print results or export to Excel,
LIMS, HTML, or PowerPoint
Search unmatched spectra
against a larger database and
validate results
Start here.
Collect and transfer data to
Spectrum Mill server
Search in variable modifications, homology,
or no-enzyme mode against the file of
validated proteins.
Check for remaining high-
quality spectra (Spectrum
Summary page)
Do remaining
spectra warrant further
processing?
Create a final results summary
(Protein/Peptide Summary page )
End here.
Basic workflow
Optional workflow
Key
No
Perform de novo sequencing (Sherenga
Sequencing and Sherenga Summary pages )
No
Additional rounds of searches
Yes
Yes
Create a file of previously validated proteins

12 Spectrum Mill Workbench Quick Start Guide
The following iterative search strategy for the Spectrum Mill workbench assumes
that the goal is to identify as many proteins as possible. If your study does not
require you to identify so many proteins, you may omit some steps (those between
step 5 and step 13).
1Copy or move the raw MS/MS data files to the Spectrum Mill server. Be sure to
set up a directory structure on the server that makes it easy to summarize and
compare your results.
2Extract the data.
3Search a database in identity mode, preferably a species subset database.
4Autovalidate the results with the highest scores, first in Protein details mode,
then (optionally) in Peptide mode. Use default settings.
5Manually review the medium-quality spectra and validate additional results (for
ion trap data down to score of 6, SPI 60 and for Agilent Q-TOF data down to
score of 5, SPI 60).
6Use Tool Belt to create a saved results file of validated protein hits (*.res file).
7Search the spectra that have not been validated. Search them in variable
modifications mode against previous validated results. Search as follows with
autovalidation and manual validation between each set of conditions:
aCombine the following common modifications into a single search: oxidized
methionine (methionine sulfoxide) and pyroglutamic acid (N-terminal).
bIf you suspect phosphorylation, combine the following modifications into a
single search: phosphorylated S, phosphorylated T, and phosphorylated Y.
cSearch for any other modifications you suspect for your sample.
8Search in no enzyme mode against the previous validated results to find
semi-tryptic or non-specific cleavages for proteins you have already identified.
(Set Digest: to No enzyme.) Repeat autovalidation and manual validation.
9When you think you are done, list sequence-not-validated spectra in Protein
Summary Details mode and look for proteins with multiple peptides. These may
represent legitimate proteins at low levels. Re-examine the spectra to confirm.
10 Optional: Search again using a larger database (entire database or larger
subset). This is most important when the original subspecies is not
well-represented in the database. Autovalidate and manually validate.
11 Check statistics in Tool Belt. If there is a significant number of unmatched
filtered spectra, continue searching.
12 Use Spectrum Summary to check for sequence-not-validated spectra with
sequence tags greater than 6 or 7. Review these and mark as Good Spectrum as
appropriate.
13 Subject the good spectra to de novo sequencing.
14 When you have gained enough information from your data, summarize the
results.

Spectrum Mill Workbench Quick Start Guide 13
Roadmap for MS-only data processing
See Chapter 6 in the Application Guide for details.
f
Set search parameters and run
searches (PMF Search page)
Summarize and review results
(PMF Summary page)
Do you want to
process the data
further?
Search another
database and review
results (PMF Search
page and PMF
Summary page)
Basic
workflow
Optional
workflow
Key
No
Yes
Prepare a final
results summary
(PMF Summary
page). End here.
Print results or
export to Excel,
LIMS, HTML, or
PowerPoint
Start here.
Prepare and transfer
mass list files to
Spectrum Mill
server
Do you have QSTAR, or
Agilent TOF or Trap data?
Preprocess
spectral files
(Data Extractor
page)
Yes
No

14 Spectrum Mill Workbench Quick Start Guide
Modes for Protein/Peptide Summary
See Chapter 2 in the Application Guide for details.
If you want to: And you want results
organized by:
Then use this mode: Example application
Validate results Peptides Peptide Manual review and validation of MS/MS search
results, organized by peptide
Proteins, then peptides Protein Summary
Details
Manual review and validation of MS/MS search
results, organized by protein
Samples, then proteins,
then peptides
Protein-Sample
Centric Rows Details
Manual review and validation of MS/MS search
results, organized by sample
Summarize results
by proteins
Proteins only Protein Summary List of all proteins identified in the data
Proteins, then samples Protein-Protein
Comparison Columns
Compare two or more samples, each of which
may contain multiple fractions. Each sample
(with all fractions) is organized in a separate
directory.
Proteins, then
redundant hits
Protein-Protein
Comparison
Redundant
Same as immediately above, with additional
detail on isoforms of proteins
Proteins, then peptides Protein Summary
Details
View proteins, with supporting peptide details
Summarize results
by samples
Samples, then proteins Protein-Sample
Centric Rows
View proteins from multiple 2D gel spots
organized in a single directory
Samples, then proteins,
then peptides
Protein-Sample
Centric Rows Details
Same as immediately above, with supporting
peptide details
Summarize results
by peptides
Peptides only Peptide List of all peptides identified in the data
Peptides, then samples Protein-Peptide
Distribution Columns
Method development (evaluation of 2D
LC/MS/MS or other fractionation scheme)
Peptides, then samples Protein-Peptide
Comparison Columns
Evaluation of fractionation scheme (provides
more information and easier export to Excel)
View a list of
proteins identified
via a single peptide
Peptides only Protein-Single
Peptide ID
Examination of results where a single peptide
was used to generate a protein identification

Spectrum Mill Workbench Quick Start Guide 15
Guide to common buttons
Any green button Initiates a process. Also saves form settings (except for data directory) until you
close your web browser window.
Save Settings Saves form settings (except for data directory) for future web browser sessions.
Select Selects a data directory for this and subsequent forms.
Mark the Make Default check box to retain this directory for future web browser
sessions.
Choose Selects peptide modifications. If you wish to view details about the modifications
that are currently displayed on your server, click the Details button on the lower
right of the Choose Modifications dialog. For more details about choosing amino acid
modifications, see the online help.
Reset Resets settings to those last saved with a green button or Save Settings.

16 Spectrum Mill Workbench Quick Start Guide
Interpreting scores from MS/MS Search
The following definitions and guidelines will help you evaluate your MS/MS search
results:
Peptide score Score of an individual peptide. This score reflects the information content
(amount of useful fragmentation) in the MS/MS spectrum.
Protein score Score of the overall protein, calculated by adding scores of all the peptides
detected for the protein
•Scores greater than 25 almost always represent valid results.
% SPI Percentage of the extracted MS/MS ion current explained by theoretical
fragmentation of the database hit
Guidelines for validating ion trap results based on MS/MS Search scores
Guidelines for validating Agilent Q-TOF results based on MS/MS Search scores
Because the mass accuracy is better than with ion trap data, you search with a
narrower mass tolerance, so there is a better chance that lower-scoring results are
valid.
When you manually validate, apply the following filters in Protein/Peptide
Summary:
•Peptide score > 5
•% SPI > 60
Peptide
score
Quality Peptide
fragmentation
Likelihood of valid interpretation
13-25 Excellent Extensive When combined with % SPI of 70 or
greater, very likely to be valid
9-13 Good Substantial When combined with % SPI of 70 or
greater, likely to be valid
6-9 Mediocre Moderate Review results to determine whether
interpretation is valid
3-6 Weak Sparse Not likely to be valid

Spectrum Mill Workbench Quick Start Guide 17

18 Spectrum Mill Workbench Quick Start Guide

Spectrum Mill Workbench Quick Start Guide 19

Agilent Technologies
© Agilent Technologies, Inc. 2007
Printed in USA
Fifth edition, June 2007
*G2721-90028*
G2721-90028
www.agilent.com
In this Book
The Quick Start Guide
presents first steps to use
the Spectrum Mill MS
Proteomics Workbench.