Introducing Kuiper: A memory and runtime efficient search-engine for unspecific directDIA analysis in Spectronaut 20

Mass spectrometry has been an important cornerstone for the study of the human immune system. While library-free DIA has become the method of choice for bottom-up proteomics, immunopeptidomics samples have remained challenging to this analysis mode. In comparison to DDA or library-based DIA, this mode of analysis faces two main challenges a) Overall lower number of identifications, and b) Significantly slower search. The underlying cause is the exponentially larger search space which negatively affects both sensitivity and speed.

While a typical fully tryptic search-space of the Human proteome (7-52 peptide length) contains around 550’000 unique peptides (without missed cleavages), an unspecific search-space of the same length range amounts to almost 1’000x this complexity or almost 500 million sequences.

Here we introduce Kuiper, a novel search-engine, integrated into the directDIA+ pipeline that was purpose built for dealing with the unique challenges of unspecific search-spaces. The integration of Kuiper into our search-engine suite allows for unprecedented capabilities for processing unspecific search-spaces. Additionally to the closed search option, Kuiper also supports an “open-search” mode (PTM Probing search) that can handle dozens of variable modifications in combination with a fully unspecific search space.