Search

Peptidi so zaporedja aminokislin. če imajo strukturni motiv obvite vijačnice, lahko predvidimo njihovo povezovanje in ustvarimo večje strukture. Za sestavljanje stabilne periodične mreže uporabljamo verige peptidov, ki jih v matematičnem modelu predstavimo kot graf verižne strukture. Tega vlagamo v ravninsko mrežo, da zagotovimo željeno obliko. Relacije med členi odločajo o tem, ali se bodo ti sami sestavili v mrežo. Matrika relacij je zaradi uporabe ortogonalne množice členov permutacijska. če upoštevamo pogoje iz narave, stabilnosti in oblike mreže, lahko možnosti še dodatno omejimo. Predlagamo omejitev območja iskanja periodične mreže ter preverjanja enoličnosti. Predstavimo algoritem z lenim vlaganjem, kjer mrežo gradimo s postopnim povezovanjem členov, ki jih vlagamo, ko jim za to preostane le ena možnost. Peptides are sequences of amino acids. In case of their coiled coil structural motif we can predict what bonds will form and create larger structures. We use chains made of peptides for assembly of stable periodic nets. We represent peptides as a chain structure graph in our mathematical model, which we embed into a planar net to guarantee the proper form. Relations between links given in a relation matrix decide whether or not links will self-assemble into a net. Because of the use of orthogonal link sets, relation matrices are permutation matrices. We can further limit relation matrices by applying several conditions derived from nature, stability and net shapes. For the search of periodic nets and uniqueness check we propose area restrictions. The presented lazy embedding algorithm connects one pair of links in a step and embeds them only if there is only one embedding option left.