While we only tested alignments between pre-computed protein family models, core regions may be inferred from the continuous regions of any protein family alignment.
In silico PT data sets can be generated from simulated or actual sequencing data and are used to test alignment through variant detection and annotation steps.
In this case report, we apply these frameworks to test alignment with practitioner experience and to examine the degree to which they contribute to the understanding of the case.
First, clustering optimization enabled us to test alignment programs and distance functions and to identify the best approach es).
In developing fold recognition methods, the SALIGN dataset [29] and Lindahl dataset [31] have been widely used to test alignment accuracy and fold recognition sensitivity, respectively.
The cranial base region was chosen to test alignment errors in the vertical direction, the forehead region for the antero-posterior direction, while the right and left zygomatic arches were chosen for the transverse direction.
To test alignment accuracy, we construct three benchmark datasets.
While it is possible to test alignments with simulated reads, this strategy is restricted by the assumption that the simulations capture the pattern of variability observed in real sequences.
A set of protein pairs was used for testing alignments and selected so that there should be some structural similarity, but little sequence similarity.
