Complete genomic sequence of the Australian south-west genotype of Sindbis virus: Comparisons with other Sindbis strains and identification of a unique deletion in the 3 '-untranslated region

Our previous studies have shown that two distinct genotypes of Sindbis (SIN) virus occur in Australia. One of these, the Oriental/Australian type, circulates throughout most of the Australian continent, whereas the recently identified south-west (SW) genetic type appears to be restricted to a distinct geographic region located in the temperate south-west of Australia. We have now determined the complete nucleotide and translated amino acid sequences of a SW isolate of SIN virus (SW6562) and performed comparative analyses with other SIN viruses at the genomic level. The genome of SW6562 is 11,569 nucleotides in length, excluding the cap nucleotide and poly (A) tail. Overall this virus differs from the prototype SIN virus (strain AR339) by 23% in nucleotide sequence and 12.5% in amino acid sequence. Partial sequences of four regions of the genome of four SW isolates were determined and compared with the corresponding sequences from a number of SIN isolates from different regions of the World. These regions are the non-structural protein (nsP3), the E2 gene, the capsid gene, and the repeated sequence elements (RSE) of the 3'UTR. These comparisons revealed that the SW SIN viruses were more closely related to South African and European strains than to other Australian isolates of SIN virus. Thus the SW genotype of SIN virus may have been introduced into this region of Australia by viremic humans or migratory birds and subsequently evolved independently in the region. The sequence data also revealed that the SW genotype contains a unique deletion in the RSE of the 3'UTR region of the genome. Previous studies have shown that deletions in this region of the SIN genome can have significant effects on virus replication in mosquito and avian cells, which may explain the restricted distribution of this genotype of SIN virus.

Learning the dynamics of embedded clauses

Recent work by Siegelmann has shown that the computational power of recurrent neural networks matches that of Turing Machines. One important implication is that complex language classes (infinite languages with embedded clauses) can be represented in neural networks. Proofs are based on a fractal encoding of states to simulate the memory and operations of stacks. In the present work, it is shown that similar stack-like dynamics can be learned in recurrent neural networks from simple sequence prediction tasks. Two main types of network solutions are found and described qualitatively as dynamical systems: damped oscillation and entangled spiraling around fixed points. The potential and limitations of each solution type are established in terms of generalization on two different context-free languages. Both solution types constitute novel stack implementations - generally in line with Siegelmann's theoretical work - which supply insights into how embedded structures of languages can be handled in analog hardware.