"Reconstructing the historical (phylogenetic) interrelationships of biological entities (e.g., taxa, species, genes) is a substantial goal of evolutionary biology not least because knowledge of these relationships facilitates the interpretation of all comparative biological data. Due to continuing improvement of sequencing techniques and computational power within the last 20 years, molecular phylogenetic analyses have expanded steadily from single gene analyses of small taxon sets up to phylogenomic analyses of many hundreds of genes and taxa. However, this is not always a simple task – particularly when the data are subject to systematic biases that can mislead the most commonly relied upon methods. In such cases, increasing the amount of data can make recovering the correct relationships harder rather than easier. The main aim of this project is to develop and evaluate new intelligent tree reconstruction algorithms based on specific split pattern analyses among multiple sequence alignments and/or morphological character patterns that will enable a significantly more reliable and accurate reconstruction of phylogenetic trees, and ultimately the Tree of Life. The main focus is on overcoming the analytical difficulties or systematic biases resulting from branch specific heterogeneities of evolutionary substitution rates leading to the widespread analytical artefact known as Long Branch Attraction (LBA). It is anticipated that the strategies developed will have broad applicability, that they will complement existing methods and will help to resolve longstanding phylogenetic problems that appear to have been confounded by LBA. By facilitating the production of more accurate phylogenetic trees, it is expected that the new algorithms will help deliver better foundations for all branches of comparative biology. The new developed reconstruction algorithms will be implemented into an existing pipeline and thoroughly tested with both simulated and empirical data."