Performing SELEX experiments in silico

Abstract

Due to the sequence-dependent nature of the elasticity of DNA, many protein-DNA complexes and other systems in which DNA molecules must be deformed have preferences for the type of DNA sequence they interact with. SELEX (Systematic Evolution of Ligands by EXponential enrichment) experiments and similar sequence selection experiments have been used extensively to examine the (indirect readout) sequence preferences of, e.g., nucleosomes (protein spools around which DNA is wound for compactification) and DNA rings. We show how recently developed computational and theoretical tools can be used to emulate such experiments in silico. Opening up this possibility comes with several benefits. First, it allows us a better understanding of our models and systems, specifically about the roles played by the simulation temperature and the selection pressure on the sequences. Second, it allows us to compare the predictions made by the model of choice with experimental results. We find agreement on important features between predictions of the rigid base-pair model and experimental results for DNA rings and interesting differences that point out open questions in the field. Finally, our simulations allow application of the SELEX methodology to systems that are experimentally difficult to realize because they come with high energetic costs and are therefore unlikely to form spontaneously, such as very short or overwound DNA rings.

The role of DNA sequence in nucleosome breathing

“…We employ the same nucleosome model as in our previous work [8,14,20,21]. The DNA is represented by the rigid bp model (RBP) [59] which treats each bp as a rigid plate, the spatial position and orientation of which are described by six (three translational and three rotational) degrees of freedom.…”

“…The sequence dependence of the model comes into play because the stiffness (K) and intrinsic shape (q 0 ) of a given bp step depend on its chemical identity. These parameters can be found in the literature [59,60], and we use the same hybrid parameterization [61] as in [8,14,20,21]. The DNA is forced into a super-helix through a set of 28 constraints that represent the 14 binding sites to the histone octamer (see fig.…”

“…Our nucleosome model, which we have tested extensively [8,10,14,20,21], reveals the sensitivity of site exposure to sequence variation. For experiments that modify the DNA sequence as part of the experimental methodology, as is for instance done in [18], this sensitivity has the potential to distort the results.…”

Shortest paths through synonymous genomes

“…This model has been widely tested against experiments, e.g., it successfully predicts relative nucleosome affinities of various sequences [22] (as measured in Refs. [18,28,29]), the rotational positioning rules of nucleosomes [22,30] (see Refs. [18,31]), translational positioning [6] (see Refs.…”

A Bottom-Up Approach for Developing Aptasensors for Abused Drugs: Biosensors in Forensics

“…Click-SELEX is also an elegant approach that relies on chemical modification of the library via click chemistry [55]. In recent years, computational approaches have also been used to increase structural complexity or to explore the sequence space [33,[56][57][58].…”