Consequently, a search for an alternative partition of local populations without a-priori assumptions was performed using SAMOVA.
Starting from the root, each dichotomy in the tree (that increases the number of clusters) generates an alternative partition that can be evaluated.
A geographically based partition of China into two was suggested in Du et al. [ 25] and an alternative partition into three seasonality based clusters was reported in Yu et al. [ 26].
However, the additive and dominance variances reported in Su et al. [ 11] result from an alternative partitioning of genetic variance and are thus not directly comparable to the classical partitioning of genetic variance [ 12].
To compare an alternative protein partition offered by a different partition in ProtoNet, we compared the partition (i.e. 'cut') that have the largest intersection with the best clusters for IPR family.
With the crosses A × B, A × C, and A × D, the best alternative partition after AB | CD would be A | BCD, in which A × C and A × D are also segregating the QTL, but A × B should also be segregating the QTL, and note that for both partitions AB | CD and A | BCD, the QTL has effect in the same direction in the A × C and A × D crosses.
On the other hand, with the crosses A × B, A × C, and B × D (which was seen to have better performance), in the only alternative partition with two crosses segregating a QTL, AD | BC, the two crosses should have QTL effects in opposite directions (the A and D alleles both result in a decrease in phenotype), and so this should be easy to distinguish from the AB | CD partition.
In this section, we first compared alternative partition algorithms and how they impact sig-mer selections in RNA-Skim and then furnish a comparison with four alternative methods on both simulated and real data.
Finally, lineage ages were estimated again by selecting the best clock and speciation process prior, under the three alternative partition schemes (P1, P2, and P3, see above).
To understand the difference, we need to consider the sign of the QTL effect in different crosses for the true partition and the best alternative partition; these are shown in Table 1.
More importantly, we show that many of those substrate cycles are not identified if alternative partition metrics are used instead of ShReD to determine the hierarchical modularity, demonstrating the value of our ShReD-based algorithm for discovering substrate cycles.
