Affinity chromatography 'must eliminate indirect processes'

Affinity chromatography - which studies samples based on known interactions between an enzyme and substrate, or a protein and receptor - faces a need to distinguish between direct and indirect processes, say researchers.

Scientists at McGill University in Quebec and Cornell University in New York explain that the direct link between, for example, an enzyme and substrate may be disguised in findings by indirect processes at work in the same sample.

Writing in Algorithms for Molecular Biology, they have devised a way to examine the results for direct and indirect affiliations, in order to cleanse the data set of indirect - and therefore unwanted - influences.

Their algorithm filters out weakly connected nodes and estimates the direct interactions in dense regions of the results network, reconstructing the desired data with high specificity and sensitivity.

Algorithms for Molecular Biology specialises in research findings and innovations in the fields of biological sequencing, phylogeny reconstruction, machine learning and structural analysis.