W.B.Langdon . 19 April 2012 2011 papers , full list
We document a parallel non-recursive beam search GPGPU FCA CbO like algorithm written in nVidia CUDA C and test it on software module dependency graphs. Despite removing repeated calculations and optimising data structures and kernels, we do not yet see major speed ups. Instead GeForce 295 GTX and Tesla C2050 report 141072 concepts (maximal rectangles, clusters) in about one second. Future improvements in graphics hardware may make GPU implementations of Galois lattices competitive.
Keywords: FCA, GPU, GPGPU, CUDA C, SBSE, software module clustering, MDG, close-by-one, artificial intelligence, arithmetic intensity
Other UCL CS technical reports
|Mendel's discovery of the essential digital nature of inheritance.||Barbara McClintock's discovery of Transposons in Maize whereby genes move not only from parent to child but also along chromosomes.||Micrograph of genetic transfer along a pilus linking two bacteria. (1959 Akiba and Ochia discovered the first interspecies gene transfer.)||2011 Mycoplasma bacteria genes are transfered between computers, including into the reference human genome DNA sequence held by GenBank and into Affymetrix GeneChips.|
The human genome sequence database contains DNA sequences very like those of mycoplasma bacteria. It appears they infect not only molecular Biology laboratories but were picked up by experimenters from contaminated samples and inserted into GenBank as if they were human. At least one mouldy EST (Expressed Sequence Tag) has transferred from public databases to commercial tools (Affymetrix HG-U133 plus 2.0 microarrays, probeset 1570561_at). We report a second example (DA466599) and suggest there is a need to clean up genomic databases but fear current tools will be inadequate to catch genes which have jumped the silicon barrier.
arXiv:1106.4192 More Mouldy Data: Another mycoplasma gene jumps the silicon barrier into the human genome.
Keywords: Human Microbiome, Horizontal gene transfer HGT, human-associated bacteria.
data directory contains results of AF241217 and DA466599 blast runs by EBI in Cambridge
[q-bio.GN] suplementary data directory (poster).
blog slashdot blogspot
Amara D. Angelica. Der Spiegel, Verseuchte Datenbanken, 04.07.2011 page 120. Genbanken von Bakterien infiziert science.orf.at, 28.06.2011. Mouldy chips New Scientist, 60 Seconds, 13 July 2011
Other UCL CS technical reports
During six months of intensive nVidia CUDA programming of an artificial intelligence (AI) application many bugs were created. We seek to pass on the software engineering lessons learnt, particularly those relevant to parallel general-purpose computation on graphics hardware GPGPU.
A reviewer said:
Summary : GPU via CUDA programming is tricky and there are many "gotchas". This paper aims to pass on some experience of developing with CUDA to those who are experimenting with more-advanced projects.
Recently a GPGPU computation intelligence application had to be redesigned to overcome performance problems. A number of software engineering lessons were learnt from this and other projects. We describe those about obtaining high performance from nVidia GPUs and practical aspects of CUDA C software development.