Achieving the lowest possible communication latency between an FPGA accelerator card and host CPU software is critical for many applications, such as high-frequency trading, in-memory storage systems, and 5G processing. However, no freely available PCIe DMA engine (DMAE) is currently optimized for low latency. To address this, we present DMA Calypte, an open-source, platform-independent, ultra-low-latency DMAE. We implemented numerous optimizations at the DMAE, software, and device driver levels to minimize communication latency over PCIe. DMA Calypte's functionality was validated on accelerator cards featuring AMD Kintex UltraScale+ and Agilex 7 F-Series chips with a PCIe Gen3 x8 interface. We achieved a round-trip time between software and FPGA of just 790 ns and a throughput of up to 35 Gbps.

Gradient boosting (GB) is an effective and widely used type of ensemble machine-learning method. The opportunity to transform the trained GB models to the hardware level represents the potential for significant acceleration of many applications and their availability as embedded systems. In this work, we have therefore developed the LGBM2VHDL tool for the automated mapping of models trained by the LightGBM library to circuits described by VHDL. Compared to existing tools, we have used an architecture that is better suited for large-scale GB models involving up to thousands of decision trees. We have further optimized the architecture using two newly proposed techniques. By applying these techniques to the tested models, the amount of memory required was significantly reduced to almost half of the original resources, and the amount of basic configurable blocks was reduced by up to 4 times on average. The developed tool is available as open-source.

MOTIVATION: Poor protein solubility hinders the production of many therapeutic and industrially useful proteins. Experimental efforts to increase solubility are plagued by low success rates and often reduce biological activity. Computational prediction of protein expressibility and solubility in Escherichia coli using only sequence information could reduce the cost of experimental studies by enabling prioritization of highly soluble proteins. RESULTS: A new tool for sequence-based prediction of soluble protein expression in E.coli, SoluProt, was created using the gradient boosting machine technique with the TargetTrack database as a training set. When evaluated against a balanced independent test set derived from the NESG database, SoluProt's accuracy of 58.5% and AUC of 0.62 exceeded those of a suite of alternative solubility prediction tools. There is also evidence that it could significantly increase the success rate of experimental protein studies. SoluProt is freely available as a standalone program and a user-friendly webserver at https://loschmidt.chemi.muni.cz/soluprot/. AVAILABILITY AND IMPLEMENTATION: https://loschmidt.chemi.muni.cz/soluprot/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: The insertion sequence elements (IS elements) represent the smallest and the most abundant mobile elements in prokaryotic genomes. It has been shown that they play a significant role in genome organization and evolution. To better understand their function in the host genome, it is desirable to have an effective detection and annotation tool. This need becomes even more crucial when considering rapid-growing genomic and metagenomic data. The existing tools for IS elements detection and annotation are usually based on comparing sequence similarity with a database of known IS families. Thus, they have limited ability to discover distant and putative novel IS elements. RESULTS: In this paper, we present digIS, a software tool based on profile hidden Markov models assembled from catalytic domains of transposases. It shows a very good performance in detecting known IS elements when tested on datasets with manually curated annotation. The main contribution of digIS is in its ability to detect distant and putative novel IS elements while maintaining a moderate level of false positives. In this category it outperforms existing tools, especially when tested on large datasets of archaeal and bacterial genomes. CONCLUSION: We provide digIS, a software tool using a novel approach based on manually curated profile hidden Markov models, which is able to detect distant and putative novel IS elements. Although digIS can find known IS elements as well, we expect it to be used primarily by scientists interested in finding novel IS elements. The tool is available at https://github.com/janka2012/digIS.

MOTIVATION: G-quadruplexes (G4s) are one of the non-B DNA structures easily observed in vitro and assumed to form in vivo. The latest experiments with G4-specific antibodies and G4-unwinding helicase mutants confirm this conjecture. These four-stranded structures have also been shown to influence a range of molecular processes in cells. As G4s are intensively studied, it is often desirable to screen DNA sequences and pinpoint the precise locations where they might form. RESULTS: We describe and have tested a newly developed Bioconductor package for identifying potential quadruplex-forming sequences (PQS). The package is easy-to-use, flexible and customizable. It allows for sequence searches that accommodate possible divergences from the optimal G4 base composition. A novel aspect of our research was the creation and training (parametrization) of an advanced scoring model which resulted in increased precision compared to similar tools. We demonstrate that the algorithm behind the searches has a 96% accuracy on 392 currently known and experimentally observed G4 structures. We also carried out searches against the recent G4-seq data to verify how well we can identify the structures detected by that technology. The correlation with pqsfinder predictions was 0.622, higher than the correlation 0.491 obtained with the second best G4Hunter. AVAILABILITY AND IMPLEMENTATION: http://bioconductor.org/packages/pqsfinder/ This paper is based on pqsfinder-1.4.1. CONTACT: [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: Upgrade and integration of triplex software into the R/Bioconductor framework. RESULTS: We combined a previously published implementation of a triplex DNA search algorithm with visualization to create a versatile R/Bioconductor package 'triplex'. The new package provides functions that can be used to search Bioconductor genomes and other DNA sequence data for occurrence of nucleotide patterns capable of forming intramolecular triplexes (H-DNA). Functions producing 2D and 3D diagrams of the identified triplexes allow instant visualization of the search results. Leveraging the power of Biostrings and GRanges classes, the results get fully integrated into the existing Bioconductor framework, allowing their passage to other Genome visualization and annotation packages, such as GenomeGraphs, rtracklayer or Gviz. AVAILABILITY: R package 'triplex' is available from Bioconductor (bioconductor.org). CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Algorithms for biological sequence analysis, such as approximate string matching or algorithms for identification of sequence patterns supporting specific structural elements, present good opportunities for hardware acceleration. Implementation of these algorithms often results in architectures based on multidimensional arrays of computing elements. Mapping effectively these computational structures on FPGAs remains one of the challenging problems. This paper focuses on a specific hardware architecture for detection of approximate tandem repeats in sequences. We show how to create a parametrized architecture model combined with an automatic technique that can determine appropriate circuit dimensions with respect to input task parameters and the target platform properties.

MOTIVATION: Current methods for identification of potential triplex-forming sequences in genomes and similar sequence sets rely primarily on detecting homopurine and homopyrimidine tracts. Procedures capable of detecting sequences supporting imperfect, but structurally feasible intramolecular triplex structures are needed for better sequence analysis. RESULTS: We modified an algorithm for detection of approximate palindromes, so as to account for the special nature of triplex DNA structures. From available literature, we conclude that approximate triplexes tolerate two classes of errors. One, analogical to mismatches in duplex DNA, involves nucleotides in triplets that do not readily form Hoogsteen bonds. The other class involves geometrically incompatible neighboring triplets hindering proper alignment of strands for optimal hydrogen bonding and stacking. We tested the statistical properties of the algorithm, as well as its correctness when confronted with known triplex sequences. The proposed algorithm satisfactorily detects sequences with intramolecular triplex-forming potential. Its complexity is directly comparable to palindrome searching. AVAILABILITY: Our implementation of the algorithm is available at http://www.fi.muni.cz/lexa/triplex as source code and a web-based search tool. The source code compiles into a library providing searching capability to other programs, as well as into a stand-alone command-line application based on this library. CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Understanding the structure and function of DNA sequences represents an important area of research in modern biology. Unfortunately, analysis of such data is often complicated by the presence of mutations introduced by evolutionary processes. At the lowest scale, these usually occur in biological sequences as character substitutions, insertions or deletions (indel). They increase the time-complexity of algorithms for sequence analysis by introducing an element of uncertainty, complicating their practical usage. One class of such algorithms has been designed to search for tandem repeats with possible errors - approximate tandem repeats. This paper investigates the possibilities for hardware acceleration of approximate tandem repeat searching and describes a parametrized architecture suitable for chips with FPGA technology. The proposed architecture is able to detect tandems with both types of errors (mismatches and indels) and does not limit the length of detected tandem. A prototype of the circuit was implemented in VHDL language and synthesized for Virtex5 technology. Application on test sequences shows that the circuit is able to speed up tandem searching in orders of thousands in comparison with the best-known software method relying on suffix arrays.

Understanding the structure and function of DNA sequences represents an important area of research in modern biology. One of the interesting structures occurring in DNA is a palindrome. Biologists believe that palindromes play an important role in regulation of gene activity and other cell processes because they are often observed near promoters, introns and specific untranslated regions. Unfortunately, the time complexity of algorithms for palindrome detection increases when mutations in the form of character insertions, deletions or substitutions are taken into consideration. In recent years, several works have been aimed at acceleration of such algorithms using dedicated circuits capable of potentially large-scale searching. However, widespread use of such circuits is often complicated by varying user task details or the need to use a specific target platform. The objective of this work is therefore to create a model of hardware architecture for approximate palindrome detection and develop a technique for automatic mapping of this model to the target platform without intervention of an experienced designer. The proposed model and the mapping technique are implemented and evaluated on a family of chips with Virtex5 technology.

The division of an application between a conventional processor and an acceleration card with FPGA chips has been proved as a suitable way for an acceleration of computationally intensive tasks. In such applications, the designer usually has to implement an interconnection between components placed in FPGA and the host system bus. This task is often complicated by different requirements of user components for throughput, latency of reading operations, need for DMA transfers etc. The objective of this work is to show a new approach for implementation of interconnection systems and to enable the designer to focus on the development of the target application. The proposed interconnection system is based on tree topology. The system eliminates the sensitivity of wide buses to the distance, supports the connection of components with different requirements for throughput, supports split transaction model and many other features. The proposed system is implemented and evaluated on chips with Virtex 5 technology.

Understanding the structure and function of DNA sequences represents an important area of research in modern biology. Unfortunately, analysis of such data is often complicated by the presence of mutations introduced by evolutionary processes. They increase the time-complexity of algorithms for sequence analysis by introducing an element of uncertainty, complicating their practical usage. One class of such algorithms has been designed to search for palindromes with possible errors-approximate palindromes. The best state-of-the-art methods implemented in software show time-complexity between linear and quadratic, depending on required input parameters. This paper investigates the possibilities for hardware acceleration of approximate palindrome searching and describes a parametrized architecture suitable for chips with FPGA technology. A prototype of the proposed architecture was implemented in VHDL language and synthesized for Virtex technology. Application on test sequences shows that the circuit is able to speed up palindrome searching by up to 8000× in comparison with the best-known software method relying on suffix arrays.

In this paper, the methodology for automated design of checker for communication protocol testing is presented. Based on the level of checking, different design strategies can be performed - in the paper the lowest level is presented. The definition of dedicated language for the description of possible communication faults is presented. The core generator is used to produce VHDL code describing the behaviour of the checker.

In this paper, we investigate three different realizations of the same block from different points of view. The mentioned different realizations include two realizations with embedded processors (custom 16-bit RISC processor and general soft-core processor) and the third realization uses Handel-C as an example of synthesisable high-level abstraction languages. The results show that development time of complete solution (HW and SW) is approximately the same for the Handel-C design and the design with soft-core processor; the development time of the Custom 16-bit RISC processor is about five times higher. Moreover, the throughput of the Handel-C design measured in the number of bits processed in one second is the highest. The obtained frequency and occupied area of the Handel-C design depends on the complexity of the used program. However, results are comparable or even better than results of the embedded processors.

Multifunctional digital circuits are circuits composed of polymorphic (multifunctional) gates. In addition to its standard logic function (such as NAND), a polymorphic gate exhibits another logic function (such as NOR) which is activated under a specific condition, for example, when Vdd, temperature or illumination reaches a certain level. This paper describes the evolutionary design of multifunctional combinational circuits at the gate level using a circuit simulator and in a field programmable gate array (FPGA). The FPGA-based implementation exhibits a significant speedup against a highly optimized software simulator.