Multisearch

Abstract: In this paper we design and analyse parallel algorithms with the goal to get exact bounds on their speed-ups on real machines. For this purpose we define an extension of Valiant's BSP model, BSP*, that rewards blockwise communication, and uses Valiant's notion of c-optimality. Intuitively a c-optimal parallel algorithm for p processors achieves speed-up close to p/c. We consider the Multisearch problem: Assume a strip in 2D to be partitioned into m segments. Given n query points in the strip, the task is to locate, for each query, its segment. For m ≤ n we present a deterministic BSP* algorithm that is 1-optimal, if n = Ω(p log2p). For m > n, we present a randomized BSP* algorithm that is (1 + δ)-optimal for arbitrary δ>0, m ≤ 2p and n=Ω(p log2p). Both results hold for a wide range of BSP* parameters where the range becomes larger with growing input sizes m and n. We further report on implementation work in progress. Previous parallel algorithms for Multisearch were far away from being c-optimal in our model and do not consider blockwise communication.

Abstract: With the proliferation of search engines for the analysis of MS data, multisearch techniques aimed at boosting the discriminating power of the search engines' score functions have recently become popular. Much statistical and algorithmic work has been done, therefore, in order to be able to combine and parse multiple search streams. However, multisearch techniques suffer from long run times, and may have little impact on false negatives because of similar peptide filtering heuristics between searches. This review focuses, rather, on multipass techniques, which use the results of one search to guide the selection of spectra, parameters and sequences in subsequent searches. This reduces the number of false-negative peptide identifications due to peptide candidate filtering while preserving statistical significance of existing (correct) identifications. Furthermore, this technique avoids substantial increases in running time and, by limiting the search space, does not reduce the statistical significance of correct identifications or introduce a statistically significant number of false-positive identifications. However, we argue that the existing combiner tools are not reliably applicable to these multipass situations, because of algorithmic assumptions about search space and statistical assumptions about the rate of true positives. Here we provide an overview of the advantages of and issues in multipass analysis techniques, the existing methods and workflows available to proteomic researchers, and the unsolved statistical and algorithmic issues amenable to future research.

Abstract: An apparatus and method manages timings of Node Bs by performing a first-step cell search on primary synchronization channel signals transmitted from the Node Bs, and stores frame boundary indexes representing frame boundaries for the Node Bs by performing a second-step cell search on secondary synchronization channel signals received from the Node Bs. Upon receiving a third-step cell search start command, the apparatus and method compares frame boundary timing corresponding to each of the frame boundary indexes to a current timing from a reference timer, performs a third-step cell search on hypotheses corresponding to the frame boundary indexes at a next slot boundary following the current timing by using scrambling codes in a Node B code group obtained by the second-step cell search, and provides a third-step cell search complete information after completing the third-step cell search on hypotheses corresponding to the frame boundary indexes.