responses varied by a 21-fold range, while in the NSI MS (Fig. 11.4b) the
responses varied only by 2.2-fold across all compounds. Thus, the metabolite
identification studies with NSI systems can potentially be semiquantitative in
nature (Hop et al., 2005; Valaskovic et al., 2006).
11.5 Software-Assisted Metabolite Identification
11.5.1 Data-Dependent Acquisition (DDA)
Detection, characterization, and identification of metabolites require multiple
mass spectrometry experiments such as full scan LC/MS and LC/MS/MS. The
process becomes very time consuming, particularly when a large number of
metabolites are formed. The interpretation of LC/MS/MS data is very
laborious and inefficient, and can be a rate-limiting step in the metabolite
identification process. Therefore, new approaches to data acquisition that
would minimize the need for multiple experiments, and data processing tools
that would simplify mass spectral interpretation, are highly desired.
2w?>Intelligent data-dependent acquisition (DDA) processes have been
developed to maximize the qualitative mass spectral information that may be
obtained from a single run. In general, a survey scan (full scan MS, precursor
ion scan, or constant neutral loss scan) is acquired as each analyte elutes from
the LC column in real-time. The data system then analyzes the mass spectra
to determine if the precursor ion should be isolated for subsequent MS/MS
experiments based on the predetermined selection criteria. If the precursor
ion meets predetermine user-defined criteria, then it automatically switches to
MS/MS mode for a predefined time. Once the product ion mass spectra are
acquired, the system returns back to the survey scan and acquires data until
the next MS/MS experiment is triggered. This approach has been widely used
for the rapid identification of drug metabolites in complex biological matrices
(Fernandez-Metzler et al., 1999; Gu and Lim, 2001; Lopez et al., 1998;
Ramanathan et al., 2002; Tiller et al., 1998; Yu et al., 1999). DDA
significantly reduces both the data acquisition and the data interpretation
process and, therefore, dramatically increases the throughput of metabolite
identification.
The data-dependent metabolite identification strategies can be configured to
incorporate the masses of expected or predicted metabolites into the mass
inclusion list. The use of an inclusion list increases the probability of acquiring
product ion MS/MS spectra for low level drug-derived materials in complex
biological matrices by forcing the trigger to occur even at low signal-to-noise
ratios. Two types of software packages, database systems (MDLI metabolite
database and Accelrys’ biotransformation database) and expert systems
(META, MetabolExpert, and METEOR) have been developed and are
commercially available for the prediction of xenobiotic metabolism (Anari and
Baillie, 2005; Langowski and Long, 2002). Anari and co-workers reported the
SOFTWARE-ASSISTED METABOLITE IDENTIFICATION 345