PatBank

PatBank performs large-scale pattern-matching searches in genomic databases to identify small non-coding RNAs (ncRNAs) and candidate microRNAs (miRNAs) while minimizing memory usage.

Key Features:

• Efficient Memory Usage: PatBank performs extensive pattern-matching operations while minimizing memory usage to enable large-scale searches.
• Block Without Error Option: Implements a "block without error" approach to ensure precise sequence matching and improve alignment accuracy.
• Integration with Experimental Data: Enables comparison of computational predictions with experimental datasets such as cDNA libraries and Northern blot results for validation.

Scientific Applications:

• Small ncRNA discovery in plant genomes: Identification of novel small ncRNAs in fully sequenced Arabidopsis thaliana and in the incomplete Nicotiana tabacum genome through pattern-matching analyses.
• Analysis of small RNA cDNA libraries: Comparison of cDNA libraries generated from small RNAs (20-30 nucleotides) against genomic sequences to prioritize expressed candidates.
• Cross-species validation: Support for experimental validation such as Northern blot analyses identifying RNAs of 21–100 nucleotides in related species including Nicotiana sylvestris.

Methodology:

Pattern-matching analysis comparing unique clone sequences against genomic data, including an implemented "block without error" matching approach; and structural prediction of candidate loci using MIRFOLD to assess stable hairpin structures relevant to miRNA genesis.