This service uses cookies. They are required for the service to function correctly. By using the service you agree to their use.
Usage data
Queue length
0
Input
Input
Objective
Target rTR
Template (= parent 5’-UTR sequence)
Mutable positions
Advanced options
Number of top sequences
Max no. of consecutive mutations
Max no. of mutable positions
The RBS Generator is a tool for the design of (RBSs). with user-defined relative translation rates (rTR). It combines predictions of the deep learning model SAPIENs with the genetic algorithm GRASP (Generation via Recursive Adaption of Sequence Populations) that allows highly efficient generative 5’-UTR design in extremely large sequence spaces.

User manual

RBS Generator is based on GRASP a genetic algorithm that efficiently searches extremely large sequence pools for variants meeting user-defined criteria. To feature RBS design, it uses rTR prediction by SAPIENs to score iteratively mutated populations of candidate 5’-UTRs until they converge to the desired rTR value (i.e. design objective). Starting from randomly generated seed sequences, populations of 50 sequences are created through in silico mutation, and mutants are propagated across generations depending on how close their predicted rTR matches the design objective. This heuristic process of “in silico evolution” is repeated until several RBS designs matching the objective are found, which are suggested to the user for experimental implementation.
Input
First, the design objective must be specified selecting one of three options: Maximize or Minimize are used to increase or decrease the rTR to the highest or lowest possible level, respectively. With Target rTR, the user must additionally specify a desired target rTR value within the entire possible rTR range (0 – 100,000 a.u.).
Second, different sequence constraints are specified that define the mutational freedom and desired restrictions, under which the subsequent optimization is taking place. Under Template, the user can specify the 5’-UTR sequence from which the optimization should start (17 nts upstream of the start codon). In most cases, this should be the sequence of the parent genetic construct (e.g. plasmid) that is to be optimized with the designed sequences. Via Mutable positions, the user can indicate which positions in the Template may be mutated by GRASP . Positions are earmarked as mutable with N (full mutational freedom) or any other degenerate base following the IUPAC code.
Example:
Template: AATATCTTAGCTAAATA[ATG]
Mutable positions: AATATNNNNNNNAAATA
→
Result: The RBS Generator will iteratively create sequence populations allowing mutations only in the underlined positions of the Template. N: A, C, T or G allowed. The [ATG] start codon must not be added to the query.
While adjusting Template and Mutable positions will suffice for most purposes, RBS Generator offers the possibility to specify further sequence constraints and search parameters via the dropdown Advanced options. Here, the user can further specify the maximum number of mutations and consecutive mutations that should appear in the target sequence relative to the template (see mouseovers for further information). Further, the Number of top sequences can be adjusted to receive between 1 and 50 (default: 10) of the top solutions (i.e. best-matching with the design objective) found across all generations.
Output
As main output, RBS Generator returns the top list of 5’-UTR designs and corresponding predicted rTRs , which most closely approximate the desired target rTR . As secondary output, the uncertainty of each prediction in percent of the full rTR-scale (i.e. 1% uncertainty corresponds to ±1,000 a.u.) is specified. This parameter can additionally help users select designs, for which the model is “most sure” about its predictions (<3% uncertainty is recommended).
Example output (*.txt):
# sequence rTR_(a.u.) uncertainty_(%)
AATATAAGGAGGAAATA 92723.224 0.063
AATATATGGAGGAAATA 87790.006 0.078
AATATGGGAGGAAAATA 86443.381 0.147
...
Additionally, a graphical output file is generated detailing the optimization process over the different generations (line plot) as well as graphical representations for the top solutions, which detail the sequences and predicted rTRs (bar plot) and the positional nucleotide composition (logo plot).
Example graphical output (*.png):
Important considerations:
Via the Mutable positions field, the user defines the size of the search space in which optimal solutions are sought. If only few positions are earmarked as mutable, the SAPIENs prediction model allows for an exhaustive search for the best sequences, which will be automatically performed for small sequence space (up to 4,096 sequences). In these cases, no heuristic optimization is needed and correspondingly the line plot above not shown. While this yields globally optimal designs, the limited mutational freedom may restrict the accessible rTR-scale range and lead to cases where the design objective cannot be approximated to a satisfactory degree. If possible, the mutational freedom should be increased in such cases to obtain improved designs.

When you use RBS Designer in your published work, please do not forget to cite:

  • Höllerer, S., Papaxanthos, L., Gumpinger, A. C., Fischer, K., Beisel, C., Borgwardt, K., Benenson, Y., & Jeschek, M. (2020). Large-scale DNA-based phenotypic recording and deep learning enable highly accurate sequence-function mapping. Nature communications, 11(1), 3551 (https://doi.org/10.1038/s41467-020-17222-4)
  • For the genetic algorithm: publication coming soon!

Usage data
Queue length
0
Summary statistics
Jobs this month
382
Jobs all time
2136
Registered users
22

Legal Disclaimer

This website is free to all users and licensed under CC BY-ND 4.0.

All tools provided on this website are for synthetic biology research and educational purposes. While efforts have been made to ensure the accuracy and reliability of the tools, they are provided "as is" without warranty of any kind, express or implied, including but not limited to warranties of merchantability, fitness for a particular purpose, or non-infringement.

By using this tool, you acknowledge and agree that: The tools’ outputs are generated based on computational models and algorithms that may have limitations or inaccuracies. The developers, contributors, and distributors of this tool shall not be held liable for any direct, indirect, incidental, consequential, or punitive damages, including but not limited to loss of data, profits, or research opportunities, arising out of or related to your use or inability to use the tool. It is your responsibility to ensure that all use of the tool and any subsequent research or application complies with applicable laws, regulations, and ethical standards. The developers of this tool are not responsible for any misuse or unlawful activity arising from its use.

Data you provide may be collected and stored solely for the purpose of delivering and improving the service, which includes generating statistical insights to enhance user experience. We are committed to safeguarding your data and will not share, sell, or otherwise distribute your information to third parties, except as required by law.