TimeLogic® Biocomputing Solutions
scalable, high-throughput systems for bioinformatics
TimeLogic, the bioinformatics brand of Active Motif, provides high-performance Biocomputing systems that combine custom designed Field Programmable Gate Array (FPGA) circuitry with optimized implementations of fundamental bioinformatics algorithms. The result is a powerful, fully supported solution for high performance Biocomputing with industry leading price / performance.
The abundance of data being generated by Next Generation sequencing systems continues to challenge informaticians and researchers alike. Institutions struggle to keep pace with this data analysis burden because sequencing throughput continues to increase at a pace faster than Moore’s Law. A variety of possible solutions exist for this problem however, of these competing technologies, TimeLogic’s FPGA-accelerated algorithms offer the best performance per dollar.
FPGA-accelerated alignment of NGS data
VelociMapper™ runs on TimeLogic's newest FPGA-based DeCypher J1 Similarity Search Engine Accelerator to provide fast, reliable results that significantly outperform software-only or GPU-accelerated alternatives.
- Faster than Burrows-Wheeler (BWT) based short-read mapping algorithms like BWA, Bowtie, or SOAPaligner/soap2.
- High sensitivity with support for up to 7 mismatches and/or indels with
no performance penalty.
- Produces sensitive Smith-Waterman gapped alignments.
- Reliable support for FASTQ and BAM formatting.
- Produces a sorted BAM file as output for convenient integration with auxiliary tools like SAMtools and GATK.
Tera-BLAST™ is the best performing implementation of the BLAST algorithm available. Our latest FPGA-based DeCypher J1 Similarity Search Engine Accelerator imbues Tera-Blast with unprecedented performance and we significantly outperform software-only or GPU-accelerated alternatives. But don't take our word for it, see for yourself.
BLAST performance is very data dependent and as a result, standardized benchmark tests are useless. Imagine, for example, how fast a job goes when you get no hits!
Instead of inferring performance, provide us with your own data and parameter settings and let us show you what we can do.
Within the field of bioinformatics, profile hidden Markov models (HMMs) are commonly used for searching protein sequence databases. Dynamic programming algorithms that utilize HMMs, such as HMMER2.0, can be computationally intensive as they compare every query symbol to every target symbol. To address this challenge, TimeLogic developed DeCypherHMM, which is an FPGA-accelerated implementation of the HMMER2.0 algorithm.
In 2011, HMMER3.0 was introduced as a replacement to HMMER2.0 and with this update, several significant algorithmic changes were introduced. Perhaps most importantly, HMMER3.0 is a heuristic algorithm, whereas its predecessor, HMMER2.0, was a dynamic programming algorithm. These algorithmic changes significantly reduce the amount of data that HMMER3.0 processes and accordingly, performance improves dramatically.
TimeLogic has chosen to emphasize heuristic algorithm performance with our latest FPGA-based DeCypher J1 Similarity Search Engine Accelerator and as a result, we are well positioned to introduce the only accelerated implementation of the HMMER3.0 algorithm. DeCypherHMM3™ will be available in early 2013.
Accuracy is what Smith-Waterman is all about. Often cited as the Gold-Standard for alignment algorithms, Smith-Waterman thoroughly interrogates all sequence space and provides the optimal alignment for a given query / target pair. The dynamic nature of this algorithm, however, has a dramatic affect on performance, rendering Smith-Waterman usage impractical for the analysis of large data sets. DeCypherSW™, our FPGA-accelerated implementation, eliminates that problem.
By moving to a larger platform for our new FPGA-based DeCypher J1 Similarity Search Engine Accelerator, we are able to add the latest FPGA chips in greater numbers. This in turn allows us to deliver unprecedented performance for the most sensitive of all alignment algorithms.
FPGA-accelerated Gene Finding algorithm
Our GeneDetective™ algorithm allows you to rapidly generate Human gene models for the study of alternative splicing events. GeneDetective™ reconstructs splice junctions in order to accurately align EST/cDNAs, proteins, or Hidden Markov Models to genomic DNA to produce a graphical gene model, multiple sequence alignment or RNA transcripts.
Request a Benchmark Test
You supply the Data and we do the rest!
TimeLogic is the only vendor of accelerated Biocomputing systems that will allow you to conduct customized performance benchmark tests before you buy, using your own data and parameter settings, so you know exactly what performance to expect. No surprises here. We will even give you the output file for your review.
If your genomics team needs additional computing resources, have TimeLogic run a large search on one of our DeCypher systems. It's the best way to evaluate what our accelerated resources can do for your throughput.