1.How do I purchase
or renew a fully licensed, and supported, version of CycDesigN 4.0?
Contact VSNi.
2. How can I learn about the type and range of designs that can be generated
in CycDesigN 4.0?
CycDesigN 4.0
comprises two design generation modules, namely CycDesigN and CycXOver. An
extensive manual is available for each module, and is included with the software.
A full description is given in each manual as to the type and range of
designs that can be generated.
3. What support do you offer if
I run into difficulties in generating designs in CycDesigN 4.0?
In first instance all queries and questions should be
forwarded to support@vsni.co.uk.
VSNi will make every effort to assist you at this point. If however your
query requires help from the developers of CycDesigN, your request will be
forwarded to them and you will be advised on your query’s progress.
4. I have an earlier version of
CycDesigN. Why should I update to this new version?
CycDesigN 4.0 comprises
a major upgrade of earlier versions of CycDesigN and CycXOver as well as the
(previously unreleased) module CycAnalysis. CycDesigN 4.0 is fully supported through
VSNi, and any new upgrades, updates, corrections and fixes will
be automatically sent to you, at no additional cost.
5. How much does it cost to
purchase CycDesigN 4.0?
See (VSNi contact for details).
6. Can CycDesigN 4.0 generate
efficient designs for use in microarray experiments?
Yes. CycDesigN 4.0 can generate block size 2 designs for
single or factorial treatments. Latinized row-column designs can be used to
adjust for dye effects. Large scale row-column designs or unreplicated
designs can also be generated for control of within-slide variation.
7. Can designs with unequal
treatment replications be constructed in CycDesigN 4.0?
Not in the CycDesigN module. However,
non-resolvable row-column designs with unequal treatment replications can be
constructed in the CycXOver module.
8. Can designs with unequal
block sizes be constructed in CycDesigN 4.0?
Resolvable block designs with equal block sizes require
that the number of treatments is a multiple of the block size. If this is
not the case then, in the CycDesigN module, a resolvable design with
unequal block sizes will be constructed automatically. Block sizes will be
kept as equal as possible. Full latinization features are also available
for such designs.
9. Should I use a resolvable
or non-resolvable design?
Resolvable designs are usually recommended for experiments with larger
treatment numbers such as in field trials. Mixed model analysis can take
advantage of resolvability by removing replicate variation orthogonal to
treatments, resulting in better recovery of inter-block information. For smaller
experiments, the resolvability condition may result in a loss of efficiency, which can outweigh the advantages of resolvability.
10. How can I analyse the
yields of a field trial that used a design generated in CycDesigN 4.0?
The post-design module
CycAnalysis provides a convenient method
for the manipulation of treatment, blocking and plot factors for incomplete block and row-column designs
generated by CycDesigN and CycXOver.
CycAnalysis also produces
GenStat and SAS code for the
mixed-model analysis of yield data
(see CycAnalysis).
11. How are efficient designs
generated?
The algorithm used to generate efficient designs is based upon an
interchange algorithm and an optimization heuristic (nested simulated annealing). Essentially we start with a design, usually
not very efficient, and search for better designs by randomly interchanging
treatments. The optimization heuristic ensures that we find designs
that reach, or nearly reach, the efficiency factor upper bound on most occasions.
Because of the random nature of the algorithm, successive runs of CycDesigN
or CycXover will not always give the same outcome. Extensive
testing of the algorithm has indicated that the designs generated are as
good, if not better, than the designs obtained from other sources. Many of
the designs generated in CycDesigN 4.0 cannot be found elsewhere.
12. For what designs will
CycAnalysis generate GenStat or SAS code?
Genstat or
SAS code is only generated for major design types
from CycDesigN (such as incomplete block, row-column, latinized, nested,
factorial), and the additive and no-carry-over models in CycXOver.
13. I cannot get the program to
run as it does not recognise my licence, what should I do?
If you have trouble in getting the program to run after you have purchased and
received a licence then you should contact
support@vsni.co.uk.
Back to Top
