Home

Introduction

Separation techniques hyphenated high-resolution mass spectrometry (X-HRMS) are key analytical instruments in a multitude of domains (e.g. proteomics, metabolomics, environmental analysis, forensic, food adulteration, quality control). Those tools allow to detect and quantify trace elements in very complex matrices. However, the data resulting from a single run can be extremely complex with files size ranging from 0.1 GB to few GB. While it is relatively easy to obtain accurate information about the main compounds or the compounds that are expected to be present (targeted analysis), much information is overlooked. Finnee is a Matlab toolbox whose aims is to provide transparent chemometrics and data mining algorithms and visualisation tools allowing a deeper analysis of data from a single run. Finnee focusses on data integrity by creating multiple datasets within a single run where a dataset contains all the information following a data transformation (e.g. filtering, centroid, baseline correction) allowing to assess in detail the effect of a particular transformation. Finnee only works with mzML files. This is now an accepted format in the MS and hyphenated MS community, and many tools can be used to transform proprietary file to mzML format; msConvert from ProteoWizard is probably the most comprehensive one.

Goals

Finnee aims to be open to researchers independently of their programming skills. Ideally, the tools will allow any analytical chemist to submit their own algorithms and compare it with alternatives as well as allow anyone to use the different approaches in their current work. Finnee differs from alternative from by its goals. Finnee is independent of the type of analyte and does not aim to compare multiple datasets. The key goal is to allow to extract precise, accurate and clear information from a given dataset. While mzML with centroid scans, we recommend using profile scans to validate the centroid algorithm.

Matlab and object-oriented programming (OOP)

Matlab is a numerical computing environment and computing language particularly adapted to work with large matrices as it in the case in this work. A working version is necessary to use the different functions; Finnee has been developed using Matlab 2016a. While no programming skills are needed to use Finnee, users should be familiar with Matlab. Very good tutorials can be found online, for example here, here (4 hours video) or here (~40 min video).

This is a second version of the Finnee toolbox; the original version can be found at Finnee. In comparison to the original Finnee, Finnee2016 used object-oriented programming (OOP) allowing higher flexibility to program and add new functionalities when needed. The toolbox consists of three main objects (Finnee, Dataset and Trace) as well as a series of secondary objects and functions. More information about object-oriented programming with Matlab can be found at MatLab

Finnee2016 is built around three key objects. Finnee is the main object that contains all information about the run. Finnee can also contain multiple Dataset objects. A Dataset contains all information from a given run. At creation, a Finnee object will only contain the original Dataset that is the succession of MS scan recorded during the separation. However, additional Dataset can be created for example after filtration, centroidisation or reconstructed via extracted ions profiles. Such approach allows keeping all information and thus check the different transformation. It should be noted that data are recorded in separated binary files. Dataset contains a series of Trace objects. A Trace is a two-dimensional representation such as MS scan or profiles. Trace will contain information (e.g. axe label) as well the name and indexation in the binary file where the information is stored.