Using genetic algorithms to select most predictive protein features

Many important characteristics of proteins such as biochemical activity and subcellular localization present a challenge to machine-learning methods: it is often difficult to encode the appropriate input features at the residue level for the purpose of making a prediction for the entire protein. The problem is usually that the biophysics of the connection between a machine-learning method's input (sequence feature) and its output (observed phenomenon to be predicted) remains unknown; in other words, we may only know that a certain protein is an enzyme (output) without knowing which region may contain the active site residues (input). The goal then becomes to dissect a protein into a vast set of sequence-derived features and to correlate those features with the desired output. We introduce a framework that begins with a set of global sequence features and then vastly expands the feature space by generically encoding the coexistence of residue-based features. It is this combination of individual features, that is the step from the fractions of serine and buried (input space 20 + 2) to the fraction of buried serine (input space 20 * 2) that implicitly shifts the search space from global feature inputs to features that can capture very local evidence such as a the individual residues of a catalytic triad. The vast feature space created is explored by a genetic algorithm (GA) paired with neural networks and support vector machines. We find that the GA is critical for selecting combinations of features that are neither too general resulting in poor performance, nor too specific, leading to overtraining. The final framework manages to effectively sample a feature space that is far too large for exhaustive enumeration. We demonstrate the power of the concept by applying it to prediction of protein enzymatic activity.     

Transfer of small chromosome fragments of<Emphasis Type="Italic">Agropyron elongatum</Emphasis> to wheat chromosome via asymmetric somatic hybridization

The chromosome constitution of hybrids and chromatin patterns of Agropyron elongatum (Host)Neviski in F₅ somatic hybrid lines -1–3 and I-1-9 between Triticum aestivum L. and A. elongatum were analyzed. Based on the statistic data of pollen mother cells, F₅ I-1-9 and-1-3 had 20–21 bivalents with a frequency of 84.66% and 85.28%, of which, 89.83% and 89.57% were ring bivalents. The result indicated that both hybrid lines were basically stable in the chromosome constitution and behavior. RAPD analysis showed that the two hybrids contained biparental and integrated DNA. GISH (Genome in situ hybridization) revealed that in the form of small chromosome segments, A. elongatum chromatin was scattered on 4–6 wheat chromosomes near by the region of centromere and telomere in the two hybrid lines. SSR analysis indicated that A. elongatum DNA segments were distributed on the 2A, 5B, 6B and 2D wheat chromosomes in the hybrids, which was in accordance with the GISH results that small-segments intercalated poly-site.     

DIFFERENCES IN SEQUENCES OF RIBOSOMAL DNA SECOND INTERNAL TRANSCRIBED SPACER AMONG THREE MEMBERS OF ANOPHELES HYRCANUS COMPLEX FROM THE REPUBLIC OF KOREA

Abstract  Differences in sequences of ribosomal DNA second internal transcribed spacer (ITS) among Anopheles sinensis, An. lesteri and An. yatsushiroensis from Korea were compared. The PCR amplified rDNA-ITS2 fragments were sequenced directly. Three samples of each species were individually determined.
Lengths of the ITS2 regions were 468bp in An sinensis , 451bp in An lesteri and 453bp in An yatsushiroenesis and GC contents were 44.87 % 46.2 % and 45.7 % respectively. Variations of the sequences ranged from 12.16 % to 30.7 % among the three species. The differences of the rDNA-ITS2 sequences would be useful for molecular identification of the three members of Anopheles hycanus complex from Korea.     

Expanded turn conformations: characterization and sequence-structure correspondence in alpha-turns with implications in helix folding

Like the beta-turns, which are characterized by a limiting distance between residues two positions apart (i, i+3), a distance criterion (involving residues at positions i and i+4) is used here to identify alpha-turns from a database of known protein structures. At least 15 classes of alpha-turns have been enumerated based on the location in the phi,psi space of the three central residues (i+1 to i+3)-one of the major being the class AAA, where the residues occupy the conventional helical backbone torsion angles. However, moving towards the C-terminal end of the turn, there is a shift in the phi,psi angles towards more negative phi, such that the electrostatic repulsion between two consecutive carbonyl oxygen atoms is reduced. Except for the last position (i+4), there is not much similarity in residue composition at different positions of hydrogen and non-hydrogen bonded AAA turns. The presence or absence of Pro at i+1 position of alpha- and beta-turns has a bearing on whether the turn is hydrogen-bonded or without a hydrogen bond. In the tertiary structure, alpha-turns are more likely to be found in beta-hairpin loops. The residue composition at the beginning of the hydrogen bonded AAA alpha-turn has similarity with type I beta-turn and N-terminal positions of helices, but the last position matches with the C-terminal capping position of helices, suggesting that the existence of a "helix cap signal" at i+4 position prevents alpha-turns from growing into helices. Our results also provide new insights into alpha-helix nucleation and folding.     

Sequence variation in the cytochrome oxidase I, internal transcribed spacer 1, and Ts14 diagnostic antigen sequences of Taenia solium isolates from South and Central America, India, and Asia.

We examined the genetic variability in the pig–human tapeworm, Taenia solium, by sequencing the genes for cytochrome oxidase I, internal transcribed spacer 1, and a diagnostic antigen, Ts14, from individual cysts isolated from Peru, Colombia, Mexico, India, China, and the Philippines. For these genes, the rate of nucleotide variation was minimal. Isolates from these countries can be distinguished based on one to eight nucleotide differences in the 396 nucleotide cytochrome oxidase I (COI) sequence. However, all of the 15 isolates from within Peru had identical COI sequences. The Ts14 sequences from India and China were identical and differed from the Peru sequence by three nucleotides in 333. These data indicate that there is minimal genetic variability within the species T. solium. Minimal variability was also seen in the ITS1 sequence, but this variation was observed within the individual. Twenty-two cloned sequences from six isolates sorted into 13 unique sequences. The variability observed within the sequences from individual cysts was as great as the variability between the isolates.     

Cloning and expression of Clostridium pasteurianum galactokinase gene in Escherichia coli K-12 and nucleotide sequence analysis of a region affecting the amount of the enzyme

The Clostridium pasteurianum galactokinase gene was cloned by complementation, of the galK locus, into Escherichia coli. Restriction enzyme analysis subcloning and Tn5 mutagenesis indicated that the gene was located on a 1.8 X 10(3) base-pair ClaI-Sau3A fragment that encoded a polypeptide of approximately 40 Mr. Although the C. pasteurianum and the E. coli galactokinases have similar subunit molecular weights, Southern hybridization analysis indicated no strong homology between their genes. Even though this clone showed a low level of galactokinase expression, the Gal+ phenotype, provided by the clostridial galactokinase, was unstable in E. coli, and the gene was frequently inactivated by the spontaneous acquisition of insertion sequences. A second clone containing this gene on a large restriction fragment was isolated by hybridization. This clone was unable to grow on galactose-containing media due to the overproduction of galactokinase. Comparison of the plasmids from these two clones revealed that the second contained an additional 300 base-pairs located at one end of the galactokinase gene. Appropriate operon fusions with a promoter-less E. coli galactokinase gene indicated that these additional 300 base-pairs had promoter activity in E. coli. The DNA sequence of this region which lies upstream of the C. pasteurianum galactokinase gene was determined and compared with that from several clones producing high, low or undetectable amounts of galactokinase. The reasons for the high and low level expression and for the instability of the C. pasteurianum galactokinase in E. coli are discussed. The presence of the galactokinase suggests that galactose is used in C. pasteurianum through the Leloir pathway via galactose 1-phosphate.     

Sequence variation and genetic diversity in the giant panda

About 336–444 bp mitochondrial D-loop region and tRNA gene were sequenced for 40 individuals of the giant panda which were collected from Mabian, Meigu, Yuexi, Baoxing, Pingwu, Qingchuan, Nanping and Baishuijiang, respectively. 9 haplotypes were found in 21 founders. The results showed that the giant panda has low genetic variations, and that there is no notable genetic isolation among geographical populations. The ancestor of the living giant panda population perhaps appeared in the late Pleistocene, and unfortunately, might have suffered bottleneck attacks. Afterwards, its genetic diversity seemed to recover to some extent. Project supported by the “8.5” Key Project of Chinese Academy of Sciences, the Chairman Foundation of Chinese Academy of Sciences, K. C. Wang Education Foundation, the Applied Basic Research Foundation of Yunnan, the National Natural Science Foundation of China, the Special Foundation for Returned Chinese Scientists, and Zoological Society of San Diego.     

Tracing the invasion characteristics of the yellow-legged hornet,Vespa velutina nigrithorax (Hymenoptera: Vespidae), in Korea using newly detected variable mitochondrial DNA sequences

The yellow-legged hornet, Vespa velutina nigrithorax (Hymenoptera: Vespidae), invaded South Korea in 2003 through Busan metropolitan city, which is located in the southeast region of the country. Previous studies aiming to trace the origin of V. velutina in Korea used a portion of mitochondrial (mt) COI and detected a single haplotype common to the site of origin. However, no subsequent study on invasive dynamics such as additional entry and/or another site of entry has been performed. In this study, segments of mt COI, CytB, and lrRNA were sequenced from 238 individuals collected in 11 Korean and two Japanese localities, but no variation in each gene was observed. Thus, we developed two intergenic spacer (IGS) sequences from the publicly available mt genome of V. velutina, which provided substantially increased variability (i.e., 19 haplotypes with 1.74% maximum sequence divergence in 1,129–1,146-bp-long concatenated sequences). Population genetic analyses using the concatenated sequences unexpectedly provided higher genetic diversity estimates in the northwest and southwest regions, both of which also harbor international cargo ports, than in the southeast region, in which Busan is located. Furthermore, this genetic result was roughly concordant with our questionnaire survey demonstrating that V. velutina was observed in apiaries located in the northwest and southwest regions up to 2012, when there was no reported prevalent distribution of the hornet beyond the southeast region. These results collectively suggest that the northwest and southwest regions of Korea are additional sites of V. velutina entry to the country, independent from the southeast region origin.