Computational approaches toward the design of pools for the in vitro selection of complex aptamers

Abstract

It is well known that using random RNA/DNA sequences for SELEX experiments will generally yield low-complexity structures. Early experimental results suggest that having a structurally diverse library, which, for instance, includes high-order junctions, may prove useful in finding new functional motifs. Here, we develop two computational methods to generate sequences that exhibit higher structural complexity and can be used to increase the overall structural diversity of initial pools for in vitro selection experiments. Random Filtering selectively increases the number of five-way junctions in RNA/DNA pools, and Genetic Filtering designs RNA/DNA pools to a specified structure distribution, whether uniform or otherwise. We show that using our computationally designed DNA pool greatly improves access to highly complex sequence structures for SELEX experiments (without losing our ability to select for common one-way and two-way junction sequences).

DNA Microarrays for Aptamer Identification and Structural Characterization

“…In contrast, a "Y" secondary structure binds all categories well (biotin-IgE, dye, and dye-IgE), indicating dye/biotin interactions may be causing the signal. It agrees with prior reports that the more complex Y structure forms more pockets for small molecule binding than a stem/loop wand structure 30 . Also, PT1 appeared in all good and poor dye and B-IgE binders, P2 bound both dye and B-IgE sequences well and contained the most G repeat sets in the structures studied, and P3 tended to be good B-IgE binders and poor dye/dye-IgE binders.…”

“…Luo et al, have developed Random Filtering and Genetic Filtering methods to increase the number of five-way junctions or to design a uniform structure distribution in the starting RNA/DNA pools 30 . One of the aptamers selected by SELEX from the designed structural libraries displayed higher affinity for ATP than a previously selected low-complexity aptamer while other aptamers showed weaker affinity.…”

Computational approaches to RNA structure prediction, analysis, and design

“…Luo et al [62] recently reported a complementary, directed-evolution-like approach for junction applications: “random filtering” to enrich 5-way junctions by repeatedly mutating junction motifs, and “genetic filtering” to produce pools with given motif distribution by a similar evolutionary approach. Similar in spirit is the patterned libraries approach by Ruff et al [63] where a specific pattern like alternating purines and pyrimidines is generated by inserting random bases between single-stranded regions.…”

Assessing the Amount of Quadruplex Structures Present within G2-Tract Synthetic Random-Sequence DNA Libraries

“…Studies of this type have been carried out computationally to investigate the types of helical structures present within random sequence libraries [34], and libraries have been designed to include desired amounts of particular helical structural motifs [35]–[37]. Assessing the level of quadruplex formation within a random sequence DNA library by computation methods is more problematic, as folding programs, such as mfold [38], were developed using the free energy of known secondary structures comprised of Watson-Crick base-pairs and do not account for competing structures such as guanine quadruplexes.…”