Download
FamLink is
compatible with all Windows versions from XP and above. Installation files are
provided in FamLinkSetup.exe below. The software can also be run in stand-alone
mode on a USB stick.
When downloading FamLink, you are encouraged to send
us an email including:
- your name
- the name and location of your institution
- any comments you might have about the program
With this information, we can keep track of how FamLink
is used, and we can notify you about possible updates of the program.
Versions of
the software may require missing vcomp100.dll which relates to some work being
done in parallel using the openmp library. A solution
is provided at https://answers.microsoft.com/en-us/windows/forum/all/windows-11-how-to-solve-vcomp100dll-was-not-found/a7f47b9d-a254-4d8c-8801-27ad50182061
and includes downloading a Windows package file for Visual C++ redistributable.
Current
version
FamLinkSetup v.2.5.2 (Upcoming release 2026-03-31)
The release has been considerably
delayed by several bugs induced by some mayor updates to the computational core
and handling of high density SNP markers (>100,000). This is a beta
version so please use it with caution, we greatly appreciate reporting of all
bugs and crashes experienced.
-
A minor bug caused the markers not to be
displayed (by default) in the database dialog
-
The theta value used in the blind search was
not saved
-
Fixed bugs when importing likelihoods from a vcf file (PL or GL tags). Using the “Enforce vcf likelihoods” now works properly
-
Blind search: searching for parent/child
relation now properly accounts for mutations (and genotype likelihoods if
activated)
Known bugs
-
Blind search with incomplete chromosome data
(e.g. data for some chromosomes is missing), does not work well the IBS
activated.
-
Activating the model for genotype likelihoods
without read or likelihood data can cause FamLink2 to crash if trying to do
calculations
FamLinkSetup
v.2.5.1 (Release 2025-08-30) Beta version 20250830
The release has been considerably
delayed by several bugs induced by some mayor updates to the computational core
and handling of high density SNP markers (>100,000). This is a beta
version so please use it with caution, we greatly appreciate reporting of all
bugs and crashes experienced.
-
The default version is now using x64 (64 bit
version) for increased memory handling. Reach out if this doesn’t work and you
need a legacy 32-bit version. We have tried FamLink2 with a set of 100,000 SNPs
markers typed in 40 individuals in the blind search interface. Keep in mind
that this generally requires large RAM (memory), say >=16 Gb.
-
Updated the computations for multiallelic markers (e.g. STRs and Microhaplotypes).
The computations should be much faster. Keep in mind that mutations are not
generally modeled by default and if turned on a more computer-intense algorithm
will be invoked.
-
Included the possibility to import genotype
data on xml format.
-
Updated the default genetic map that is
supplied with FamLink2 (credit to Lucinda Davenport for providing this),
adjusting some STR positions.
-
Included features in the “Generate data”
function to simulate dropout of markers and alleles.
-
Included function in “Blind search” dialog to
perform dynamic marker selection based on centimorgan
(cM) distance. This is particularly useful if you have dense marker data (say
>100,000 SNP markers) with different overlap for a set of samples and would
like to dynamically select markers for LR calculations for each pair of samples
in comparison.
-
Several minor updates
-
We recommend all users to reach out to us if
they intend to use the model for genotype likelihoods.
-
Updated the parameter that deals with new
alleles. There is now a separate parameter to deal
with unobserved data in the genotype likelihood model, see also https://famlink.se/f_genotypelikelihoods.html.
This parameter is unrelated to the frequency that new alleles are assigned.
-
Updated the About
dialog to properly display the correct software version (was previously stuck
at 2.3).
-
Updated the “How to cite FamLink2” dialog
with new references
-
Added “Direct match” as a possible
relationship in the blind search feature
-
Several minor updates to the blind search
dialog
-
Several updates to the genotype likelihood
model
-
Corrected numerous minor bugs
-
Appropriate handling of STR markers, previous
versions may produce incorrect results for these markers. More efficient
handling of multi-allelic markers is still on the TODO list.
-
Validation manuscript in revision
FamLinkSetup v.2.4
(Release 2023-08-24) containing both a single core and a multicore
(mc) version
-
Fixed several bugs related to Created
pedigrees. Still use created pedigrees with caution as this feature is still
slightly unstable from time to time.
-
Increased speed in several dialogs
-
Improved and corrected the Report dialog. By
default genotypes and frequency data is not displayed in the report.
-
Improved the model for genotype likelihoods.
Several improvements compared to version 2.3. Two parameters, sequence/mapping
error (default 0.001) as well as sample specific error (default 10) which
roughly corresponds to the number of original DNA molecules. Future versions
will allow setting this per sample instead of an overall value.
-
Slight updates to output from Generate data
tool.
-
Fixed and updated several minor bugs
-
Blind search and DVI search now allows to
view individual marker LRs
-
Several new settings in the Advanced dialog
-
Windowed kinship/segment approach as proposed
by Snedecor
et al. is now available in the blind search. According to the authors it
performs better for medium dense marker panels (in comparison to a traditional
segment approach)
Known
bugs/constrains
-
The “Generate data” feature can be used to
generate DNA data for pairs of relatives (or unrelated individuals). The tools allows some advanced features, e.g. admixed founders,
which has not been tested yet. However basic functionality is tested and should
work. Please play around with it!
-
Users should be observant for any bugs and
report findings.
-
FamLink
can generate debug files through activating the debug mode, File -> Settings
-> Debug mode. This will force FamLink to output
debug files to the install directory of the software that in turn can be
supplied to the developers.
Examples
Earlier
versions
-
Fixed mayor issue that would allow users to start
the software due to a missing dll file, but also
removed the feature to use multiple cores. Hopefully we will find a way back to
this feature as it can greatly increase speed.
-
Mayor updates to several features and
functions.
1. Generate
data. Found through Tools ->
Generate data and described in the tutorial. Can be used to generate
data for pairs of individuals. Compute IBS/Segment and LR statistics.
2. Phenotype/Haplogroup/Ancestry inference tools.
Found through Tools ->
Phenotype/Haplogroup/Ancestry.
Used to infer phenotypes (now eye and hair color) using published logistic
regression models with betas. Also used to infer Y-haplogroups as well as biogeographic ancestry.
3. Blind
search. Found through Tools ->
Blind search. Used to blindly search for relatives in a list of
individuals or to find the most likely relationship(s) for two (or more)
individuals. Can also be used in a DVI operation.
4. DVI search. Described in the tutorial.
Allows the user to search a list of PM profiles against a
reference family to find the most likely fit (or no fit).
5. Use
genotype likelihood in the calculations. A comprehensive
implementation of a model for sequence data. The model is activated
through File->Settings.
In addition, the model needs read data, either through a vcf
input file or through a CLC-like input file. Note that this feature is not
fully tested and will be tested in the next version of FamLink2. USE WITH CARE!
-
Fixed mayor issue that would allow users to
start the software due to a missing dll file, but
also removed the feature to use multiple cores. Hopefully we will find a way
back to this feature as it can greatly increase speed.
-
Updates to the Generate data feature. Still in Beta testing. This feature allows the user to
generate data for some given relationships, i.e. simulate data. The likelihood
computations do not currently work.
-
New added core computational algorithm
described in Mostad et al. 2022 [Manuscript in preparation].
-
The software now handles any number of linked
markers.
-
Completely new feature to model genotype
uncertainty based on sequence reads, described in Mostad et al. 2022 [Manuscript
in preparation]. Use this feature with pre-caution as the is
has not be widely tested yet.
-
The software can model population
substructure through the fixation parameter (Fst/Theta).
-
The implementation is general, meaning that
any pedigrees are allowed using any number of individuals and any number of
genetic markers.
-
Blind search feature which includes
possibility to search a set of individuals in an all-against-all search. The
feature includes possibilities to compute a) LR with different settings, b)
Segment sharing with different settings and c) Kinship coefficients.
-
Phenotype/ancestry and Y-haplotype
predictions can be performed in a special function. FamLink
can also prepare files for upload to Hirisplex and Genogeographer web sites.
-
Some computations can now be performed in
parallel such that modern computers, with several CPU cores, will be able to
complete computations in shorter time.
-
Please keep in mind that this is a Beta
version and could contain several bugs. In particular, the simulation interface
is not thoroughly tested yet.
-
Completely new interface and new computation
algorithm.
-
The software now handles any number of linked
markers.
-
Some of the main workflow is still very
similar to older versions and functions like Quick analysis (analyze Familias
files) still remain.
-
Please keep in mind that this is a Beta
version and could contain several bugs. In particular, the simulation interface
is not thoroughly tested yet.
-
Bugs corrected mainly in the Quick analyses
interface.
-
Several minor bugs corrected.
-
The file markerInfo.ini, located in the FamLink install directory can be edited to include more
predefined markers with genetic positions to be used in the Quick analysis
interface.
-
Bugs corrected in the Quick analyses
interface.
-
Several minor bugs smashed
-
Minor updates
FamLinkSetup
v.1.0
You may send comments to daniel.kling@rmv.se