@article{fdi:010068895,
  title = {{S}pectral consensus strategy for accurate reconstruction of large biological networks},
  author = {{A}ffeldt, {S}. and {S}okolovska, {N}. and {P}rifti, {E}. and {Z}ucker, {J}ean-{D}aniel},
  editor = {},
  language = {{ENG}},
  abstract = {{B}ackground: {T}he last decades witnessed an explosion of large-scale biological datasets whose analyses require the continuous development of innovative algorithms. {M}any of these high-dimensional datasets are related to large biological networks with few or no experimentally proven interactions. {A} striking example lies in the recent gut bacterial studies that provided researchers with a plethora of information sources. {D}espite a deeper knowledge of microbiome composition, inferring bacterial interactions remains a critical step that encounters significant issues, due in particular to high-dimensional settings, unknown gut bacterial taxa and unavoidable noise in sparse datasets. {S}uch data type make any a priori choice of a learning method particularly difficult and urge the need for the development of new scalable approaches. {R}esults: {W}e propose a consensus method based on spectral decomposition, named {S}pectral {C}onsensus {S}trategy, to reconstruct large networks from high-dimensional datasets. {T}his novel unsupervised approach can be applied to a broad range of biological networks and the associated spectral framework provides scalability to diverse reconstruction methods. {T}he results obtained on benchmark datasets demonstrate the interest of our approach for high-dimensional cases. {A}s a suitable example, we considered the human gut microbiome co-presence network. {F}or this application, our method successfully retrieves biologically relevant relationships and gives new insights into the topology of this complex ecosystem. {C}onclusions: {T}he {S}pectral {C}onsensus {S}trategy improves prediction precision and allows scalability of various reconstruction methods to large networks. {T}he integration of multiple reconstruction algorithms turns our approach into a robust learning method. {A}ll together, this strategy increases the confidence of predicted interactions from high-dimensional datasets without demanding computations.},
  keywords = {{N}etwork reconstruction ; {C}ommunity-based method ; {S}pectral theory ; {H}igh-dimensional data ; {M}icrobiota},
  booktitle = {},
  journal = {{BMC} {B}ioinformatics},
  volume = {17},
  numero = {16},
  pages = {art. 493 [13 p.]},
  ISSN = {1471-2105},
  year = {2016},
  DOI = {10.1186/s12859-016-1308-y},
  URL = {https://www.documentation.ird.fr/hor/fdi:010068895},
}