3D reconstruction schemes.

<p>The figure depicts the 4 reconstruction schemes employed in the study. <b>(A)</b> Sparse Hi-C models were generated by converting Hi-C contact maps to spatial nanometric distances and uniformly sampling from this map according to the desired sparseness. The non-linear program solved by Duan <i>et al</i>. [<a href="http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1004298#pcbi.1004298.ref008" target="_blank">8</a>] was utilized to generate the 3D model. Finally, the quality of the resultant models was assessed in a series of tests based on previously published results on yeast genomic organization. <b>(B)</b> Orthologous-integrated models were generated by converting the <i>S</i>. <i>pombe</i> (SP) Hi-C contact map to spatial nanometric distances, projecting it on the <i>S</i>. <i>cerevisiae</i> (SC) genome through orthologous genes and normalizing it according to SC distances. Integration into the Hi-C based map of distances was done by sampling non-overlapping (unknown) distances. <b>(C)</b> Gene interaction-integrated models were generated by utilizing the codon usage frequency similarity (CUFS) to predict the functional distance between genes. Distances were normalized according to <i>S</i>. <i>cerevisiae</i> distances, and non-overlapping distances were sampled and integrated into the model. <b>(D)</b> Random models were generated by shuffling the coordinates of the Hi-C distance map and integrating them into the model in the same manner.