PID2022-137275NA-I00

Description

This repository contains data associated with the X-BIO project: Cross-Architecture and Energy-Efficient Parallel Metaheuristics for Bioinformatics (PID2022-137275NA-I00), funded by MICIU/AEI/10.13039/501100011033 and by ERDF, EU. Data has been organized in the following folders: * CrossSustainability contains PHYLIP datasets used for the evaluation of cross-architecture models and sustainability aspects when running phylogenetic algorithms. Two Excel files (Profiling22cXeon6238.xlsx, Profiling32cEpyc7543.xlsx) are provided with energy results from two different platforms, reporting execution time (column "Time(s)"), power (column "Power(W)"), energy (column "Energy(J)"), and energy-delay product (column "EDP(J*s")). An additional file (ProfilingAccelerator.xlsx) reports profiling data from hardware accelerators, comparing time results from cross-architecture models. * Distances contains distance matrices in PHYLIP format, where the character '.' is employed to denote a missing distance between two organisms. Two Excel files containing profiling data on two different systems are provided. These Excel files contain measurements of execution time (in seconds) required by different imputation methods (column "Method") under different percentage of missing distances (from "5%" to "60%", with each percentage being in a separate column). The last column "Average" provides the average time across percentages. * Microbiome contains CSV data for microbiome research. OTU tables are given by the YB-GG-data.csv file, with each column denoting a separated OTU and each row a different sample. The YB-GG-labels.csv provides the labels for each OTU, using two columns (SampleID and label) to make the association. * MSA contains multiple sequence alignments in PHYLIP format, where the character '?' is employed to denote unknown site values. An Excel file containing profiling data under serial and parallel versions of different algorithms is provided. The upper half of the table in the file reports average Normalized Robinson-Foulds for the serial (column "NRF-Serial version") and parallel (column "NRF-Parallel version") implementations on each dataset. The lower half of the table reports execution times (in seconds) for each version (columns "Time-Serial version" and "Time-Parallel version"). * Phylogenetic contains reference sequences (PHYLIP format), reference phylogenies (Newick format) and query sequences (PHYLIP) for phylogenetic placement and evaluation. An Excel file containing profiling data for different search implementations is provided. The first column reports the name of the dataset and its experimental scenarios (i.e., number of query sequences). The remaining columns report the execution time (in seconds) for different independent runs (columns "Execution"), the median time (column "Median"), the variation coefficient (column "Var. Coef."), speedups over the serial version (column "Speedup") and parallel efficiency (column "Efficiency").

Institutions

• Universidad de Extremadura

  Extremadura, Badajoz

Categories

Funders

• Agencia Estatal de Investigación

  Ministerio de Ciencia, Innovación y Universidades

  Madrid, Madrid

  Grant ID: PID2022-137275NA-I00

Licence