Genome scale oligo design is a very computation intensive process and requires optimization over a multi-parameter space. Different genomes vary in their sequence compositions and further, nucleotide composition within and across all known gene sequences of a species also varies considerably. This makes it hard to pin down appropriate parameter choices a priori; indeed a priori estimates could well turn out to provide oligos for only a subset of the genes at hand, necessitating a relaxation in the parameters for the remaining genes. This in turn requires time-consuming recomputation of probe properties with these relaxed parameters, possibly several times, until the right parameter choices are arrived upon.

To avoid frequent oligo computation and to make the selection process efficient, Sarani uses a novel Store-and-Query approach. During the initial oligo design process, only a few broad parameters need to be specified, e.g., the desired range of lengths for the oligonucleotide probes. Sarani then computes all possible oligonucleotide probes for all gene sequences along with their properties and stores these in a relational database. Though this is a time consuming step, it needs to be performed only once. After the oligo database is built, the user can interactively query the database, view oligo probe properties and select good quality probes for specific experimental conditions.

  Iteratively Relaxed Querying

Once the oligo database is computed, thresholds can be specified for various oligonucleotide probe properties (e.g. Minimum Tm, Maximum Non-specific or Cross-Tm, GC content, Low Complexity Range etc) and the oligo database can then be queried to retrieve optimal probes. The Graphical User Interface (GUI) of Sarani allows easy selection of parameter thresholds and database querying. Using the Preview feature, the percentage of target genes having probes satisfying the chosen parameter settings can be quickly determined. When very stringent parameter settings are used, a fraction of the genes may not contain any probes that satisfy these settings. The Iterative Query feature of Sarani allows relaxation of these parameters and reruns the query on only genes which were not satisfied in previous queries. To determine which poarameters to relax, Sarani provides a way of viewing probe properties which allows quick zeroing down on the blocking parameters, i.e., those parameters which when relaxed cause a larger fraction of genes to be satisfied. This process can be iterated several times until the best possible oligonucleotide probes can be retrieved for all the genes.