Microarray time-course data relates to the recorded activity of large numbers of genes (~8,000) recorded in parallel over a number of time points (~20). This data contains a massive amount of potentially valuable information. In order to unlock that information, a biologist needs to find patterns of changing activity within variable (normally large) proportions of the data. Established techniques are, however, limited in the quantity and range of valuable patterns that they can allow a biologist to find. For example they cannot uncover patterns such as those where smaller numbers of genes have common activity over an interval of the data. This research presents an alternative approach where an animated display allows biologists to find these potentially valuable patterns. The type of animated visualisation developed is unique in that it both presents abstract data which has no spatial attributes and maps time in the data to time in the display (to animate across time). This animation is formed by presenting an interval of the data and allowing that interval to be progressively re-specified with the display updated. Here, the perception of spatial motion in the display can be related to changes in the data. This allows the user to pre-attentively perceive patterns in the data and, as the display changes over time, perceive a greater variety of patterns that may be of relevance to their analysis. The main issue involved in developing the animated visualisation was the need to configure an effective display for frames of an animation using abstract qualities. This had to be effective, revealing enough information for patterns to be detected, and expressive so that meaning could be derived from motion and patterns could be properly interpreted. In addition to this, the interface needed to accommodate the inability of humans to absorb information when it is only presented for a brief time. These issues are dealt with by matching the user's conceptualisation of changes in the data to motion in the display, giving the user direct control over the pace and direction of the animation, interpolating the data for a smooth animation, and coordinating animated/static views. This work has been funded by a Scottish Enterprise Proof of Concept grant MATSE.
Microarray Time-series Explorer (MaTSE) is a software application
developed to improve the analysis of microarray data by allowing the user to
explore their data using a unique visual interface. Throughout
the project we focused on engaging potential end-users in academia and
industry toward developing functionality relevant to their needs and producing
software that is of a standard that would attract the interest of potential
investors or licensors. In conjunction
with this we worked toward amassing commercial know-how across the project
team and building a commercialisation plan that can be used to exploit the
software and further guide development.
Consultation with our collaborating biologists has
allowed us to define three main unique selling points for the MaTSE software.
ability to easily discover patterns in data not possible with other
visualisation tools: The primary USP of MaTSE
is that it allows users to explore their data in a way that allows them to find
patterns of correlated gene activity occur over shorter intervals of time in
the data. These patterns often relate to biological phenomena of genuine
interest to biologists and they cannot be found using more traditional analysis
techniques. MaTSE also allows users to find more dominant trends in recorded
gene activity and allows users to cross reference their findings with stored
feedback and easy interpretation of results: The MaTSE software displays
microarray data without any type of complex algorithmic preprocessing such as
that applied by traditional analysis techniques such as clustering, principle
component analysis or self organising maps. While traditional techniques
attempt to summarise the data by displaying an abstraction of that data, the
only processing of the data involved in MaTSE is the calculation of
straight-forward mean and change values in the scatter-plot. This makes the
display closer to the actual recorded data and allows users to gain a better
understanding of their data during analysis without having to take time to comprehend
algorithms which might have applied to the data.
ability to store and recall patterns: The unique
scatter-plot layout in MaTSE is such that user selections are measurable (based
on activity, change in activity and mean activity) and are meaningful when
recalled. These selections, and combinations of selections, are recorded so
they can be stored, annotated and shared with other users.
MATSE: Information Visualisation of Microarray Time-course Data is a Proof of Concept Fund project funded by Scottish Enterprise.
Carried out in collaboration with and others.
For further information please refer to http://www.matse.org.uk.
Each query is stored in a list in the pattern browser component. Superfluous selections are removed after a query that replaces the previous result Brower can be used to restore and adjust parameters with continuous feedback of results
Genes are segments of DNA that carry the genetic or inherited information within all living organisms and interact with each other to influence the organism's physical development and behavior. Gene expression is a key indicator of gene activity and can be measured in microarray experiments.
Biology is a visually grounded scientific discipline-from the way data is collected and analysed to the manner in which the results are communicated to others. Traditionally pictures in scientific publications were hand-drawn; however they are now almost exclusively computer-generated.
Scottish Woodlands Ltd is Scotland's leading independent full-service forest management company. It is also the second largest operator in the UK, with a network of 17 offices operating throughout Scotland, England, Wales and Northern Ireland.
Information Management and Presentation for Research Activity and Related Data. A linked intranet and internet solution for the managing and presentation of data concerning the research related activties of university researchers.
Future energy management for buildings.
Build a replacement system for managing purchase approvals and authorised signatories within a large complex organisation.
Information visualisation is the use of enhanced Graphical User Interfaces (GUIs) to communicate and interact with complex data sets such as social networks, multiattribute tables, and financial data. Pages of text and numbers are not the most effective way to communicate or understand information.
+44 131 455 2437
Dean of Research and Innovation
+44 131 455 2772
Dr Paul Craig
(not currently an institute member)
Senior Research Fellow
+44 131 455 2738
Kennedy, J. (2013). MaTSE: the gene expression time-series explorer. BMC Bioinformatics, 14(Suppl 19), (S1), .
Kennedy, J. (2012). MaTSE: The Microarray Time-Series Explorer. In: Roerdink, J.,
Hibbs, M. (Eds.) Proceedings of IEEE Symposium on Biological Data Visualization (BioVis), 2012 , , () ( ed.). (pp. 41-48). : . IEEE.
Cannon, A. (2010). Pattern Browsing and Query Adjustment for the Exploratory Analysis and Cooperative Visualisation of Microarray Time-course Data. In: Luo, Y. (Ed.) Proceedings of the 7th International Conference on Cooperative Design, Visualisation and Engineering, 6240/2010, () ( ed.). (pp. 199-206). Mallorca, Spain: . Lecture Notes in Computer Science.
Cumming, A. (2005). Animated Interval Scatter-plot Views for the Exploratory Analysis of Large Scale Microarray Time-course Data. Information Visualization, 4, (3), 149-163.
Cumming, A. (2005). Coordinated Parallel Views for the Exploratory Analysis of Microarray Time-course Data. In: (Ed.) Proceedings of 3rd International Conference on Coordinated & Multiple Views in Exploratory Visualization, , () ( ed.). (pp. 3-14). London: . IEEE Computer Society Press.
Cumming, A. (2005). Time-series Explorer: An Animated Information Visualisation for Microarray Time-course Data. BMC Bioinformatics 2005, 6, (3), P8.
See all publications