Transcript Linkage analysis

6
Linkage analysis
Jan Hellemans
Finding causal mutations
 2 opposing strategies
 sequence then select
 select then sequence
 Sequencing
 traditional Sanger sequencing only possible after selection
 Massively parallel sequencing possible prior to or after selection
 RNA sequencing
 exome sequencing
 genome sequencing
Finding causal mutations
 Selection
 positional (prior to sequencing)
 linkage analysis
 GWAS
 structural variations (e.g. microdeletions)
 functional (prior to & after sequencing)
 candidate genes selected based on known function or involvement
in related disorders
 filtering of variants based on functional predictions
 overlap (after sequencing)
 looking for genes / variants that occur in multiple independent
patients
 mostly a combination is used
exome sequencing
Aims





Interprete microsatellite results
Add genotypes to pedigrees
Create pedigree and genotype files
Calculate and interprete LOD-scores
Delineate linkage intervals




Basic principles of linkage analysis
Analyze other types of markers
Association studies
Learn how to work with specific pedigree programs
Starting linkage analysis
Preparations
 Clearly define the phenotype
 If not specific enough than you may analyze different disorders that can
map to different genomic loci
 LOD scores are additive
 Find suitable families
 larger is better
 more patients is better
 Collect genomic DNA from as much family members as
possible
 Determine the type of inheritance
 Calculate the power to prove linkage with the available
material (SLink – not part of this course)
Linkage analysis types
 Directed linkage analysis
 Evaluate linkage at a specific locus such as a candidate gene
 Common approach: evaluate an intragenic, 5’ and 3’ marker
often microsattelites
 Genome wide linkage analysis
 Screen for linkage for markers spread across the entire genome
 Microsatellites: ~400 markers spaced at about 10cM
 SNP’s: 500k SNP array
 Homozygosity mapping
 Screen only affected individuals in inbred families
 Select homozygous markers (typically SNP markers)
 Very efficient technology
 Fine mapping
 Some linked markers are known, but the borders of the linkage interval
still need to be defined
Exercise – Part 1
 2 inbred families with a recessive disorder
 With a homozygosity mapping based on 500k SNP
arrays 2 candidate regions could be identified
 Chromosome 4
 Patient 1 homozygous for
 6.052Mb - 14.488Mb
 21.008Mb – 37.477Mb
 Patient 2 homozygous for
 11.186Mb – 37.219Mb
40,000
35,000
30,000
25,000
20,000
15,000
 Task: find microsatellite markers to
confirm linkage
10,000
5,000
1
2
Find additional flanking markers




Find physical position of marker in NCBI > UniSTS
NCBI map viewer: http://www.ncbi.nlm.nih.gov/mapview/
Go to Homo sapiens and to the wright chromosome
Maps & options: show
 DeCode, Généthon & Marshfield (genetic maps)
 Genes




Set region: e.g. 2Mb up- and downstream of your marker
Click ‘Data as table view’
Click on STS behind a marker to see its details
Select markers that
 locate to only 1 genomic location
 have a PCR product with an extended size range
one size  not polymorphic
http://www.ncbi.nlm.nih.gov/projects/mapview
http://www.ncbi.nlm.nih.gov/projects/mapview
http://www.ncbi.nlm.nih.gov/projects/mapview
Exercise – Part 1 > possible solution
 Markers in 1st candidate region





D4S3017 (21.078Mb)
D4S3044 (25.189Mb)
D4S1618 (33.857Mb)
D4S3350 (33.857Mb)
D4S2988 (36.889Mb)
 Markers in 2nd candidate region





D4S1582 (10.311Mb)
D4S2906 (12.321Mb)
D4S2944 (13.141Mb)
D4S1602 (14.059Mb)
D4S2960 (15.437Mb)
  Order primers & analyze them on all family members
Analyzing microsatellite data
Microsatellites > basics
 Repeats of short sequences (e.g. 2bp)
NNNNAC(AC)nACNNNN
 Number of repeats is variable (instable sequence)
 Number of repeats determines the allele
 Number of repeats corresponds to specific length of
PCR product:




allel 1: NNNNACACACACACNNNN
allel 2: NNNNACACACACACACNNNN
allel 3: NNNNACACACACACACACNNNN
...
(5*AC  18bp)
(6*AC  20bp)
(7*AC  22bp)
 Determine length to know the allele (sequencer)
Microsatellites > basics
Microsatellites > determine size
 Use internal size standard (other color)
220bp
230bp
225bp
Microsatellites > heterozygotes
220bp
230bp
223bp
225bp
Microsatellites > stutter peaks
 Repeats are difficult to copy  polymerase slips
 Some amplicons have 1 repeat less
a few even loose multiple repeats
 Small repeats are more prone to slippage and show
more pronounced stutter peaks
 Largest product is the correct one
 Distance between peaks = length of a repeat
Microsatellites > stutter peaks
allelic peak
1st stutter peak
2nd stutter peak
Microsatellites > stutter peaks
 Allelic peaks are the heighest
 Stutter peaks are lower
A1
A2
Microsatellites > stutter peaks
A1
A2
Microsatellites > +A peaks
 Taq polymerase tends to add an extra A at the 3’ end
 Variable degree of products with or without this extra A
 Do not confuse with stutter peaks (only 1bp difference)
allelic peak
allelic peak + A
1st stutter peak
1st stutter peak + A
2nd stutter peak
2nd stutter peak + A
Microsatellites > complex plots (stutter & +A)
A1
A2
Microsatellites > mutliplex
 Combine multiple markers in a single analysis ($$$)
 Different size range
 Multicolor
 Commercial kits: e.g. 16 markers / lane
Microsatellite plots examples
Genotyping pedigrees
Genotyping pedigrees
 Screen one or multiple markers for some or all family
members
 For every marker:
 Make a list of all occuring allele sizes
 Due to technical variation on sizing the same allele can have a slightly
different size in different measurements (-0.4bp _ +0.4bp). Give all
alleles within this range the same allele number
 Add the allele numbers to the pedigree at the corresponding
individual/marker combination
 Find the wright phase
 Advanced software like GeneMapper can generate
tables with allele numbers for every sample / marker
 Advanced pedigree programs like Progeny can store
genotype information for family members
 Verify inheritance
Exercise – Part 2
 Genotype 3 markers in all available individuals of 2
families
 Pedigrees & microsatellite plots in
ExercisePart2-GenotypingData.pdf
 Add allele numbers for the 3 markers to the pedigree
 Interprete the genotyped pedigrees: linked?
Family 1
Family 2
Exercise – Part 2 > Conclusions
 D4S1582
 Mendelian error  can not be interpreted
 D4S2944
 Linked
 D4S3017
 Not-linked: unaffected individuals with the same genotype as a patient
Calculate LOD scores
EasyLinkage
 EasyLinkage = UI for linkage analysis
 http://genetik.charite.de/hoffmann/easyLINKAGE/index.html#start
 Bioinformatics. 2005 Feb 1;21(3):405-7
PMID: 15347576
 Bioinformatics. 2005 Sep 1;21(17):3565-7
PMID: 16014370
 Interface for many linkage analysis programs
 Input




Pedigree file (linkage format)
Genotype file(s)
Marker information (already provided for popular markers)
Settings
Pedigree file
 Naming requirements for EasyLinkage:
p_xxx.pro  e.g. p_SMMD.pro
 Format:
 Tab delimited text file
 1 individual per row
 Columns:








1  family ID
2  person ID
3  father ID
4  mother ID
5  sex (1=male, 2=female, 0=unknown)
6  affection status (1=unaffected, 2=affected, 0=unknown)
7  DNA availability (optional, relevant for power calculations)
8  liability class (to be provided if multiple liability classes are used)
Genotype files
 Person ID’s have to match exactly with those provided in
the pedigree file
 Naming requirements for EasyLinkage:
MarkerName_xxx.abi  e.g. D1S1609_SMMD.abi
 Format:
 Tab delimited text file
 1 individual per row
 Columns (for microsatellite based analysis):
 1  marker (same as in file name and matching a marker in an
available marker set)
 2  custom information (content doesn’t matter, but column must be
present)
 3  individual ID (match person ID in pedigree file)
 4 & 5  genotypes for 2 alleles (unknown=0)
Marker information
 Contains information on the chromosome and position of
every marker
 Already available for a number of commercial SNParrays and for the microsatellite markers from
 Genethon
 Marshfield
 DeCode
 Custom marker sets can be created (see manual)
EasyLinkage settings
 Choose a program:











FastLink  Parametric, single-point
SuperLink  Parametric, single-/multipoint
SPLink  Nonparametric, single-point
Genehunter  Nonpara-/parametric, single-/multipoint
Genehunter Plus  Nonpara-/parametric, single-/multipoint
Genehunter MOD  Nonpara-/parametric, single-/multipoint
Genehunter Imprinting  Nonpara-/parametric, single-/multipoint
GeneHunter TwoLocus  Parametric, two-locus, single-/multipoint
Merlin  Nonpara-/parametric, single-/multipoint
SimWalk  Nonparametric, single-/multipoint
Allegro  Nonpara-/parametric, single-/multipoint & simulation, single/multi-point
 PedCheck  Mendelian error check
 FastSLink  Simulation, single-/multi-point
EasyLinkage settings




Parametric <-> non-parametric
Single point <-> multipoint
Frequency of the disease allele
Penetrance vectors (wt/wt, wt/mt, mt/mt)





Standard dominant: 0 1 1
Standard recessive: 0 0 1
Reduced penetrance: replace 1 by penetrance (e.g. 0.9)
Phenocopy: replace 0 by percentage of phenocopy (e.g. 0.1)
Example: 0.01 0.9 0.99
1% chance to show a similar phenotype despite a normal genotype
90% chance to show the phenotype when 1 mutant allele (dominant
with incomplete penetrance)
99% likelihood to present with the phenotype if both alleles are mutant
Evaluate calculated LOD-scores
 Maximum LOD-scores can be seen in EasyLinkage
 Details about LOD-scores at different recombination
fractions can be found in text files generated by
EasyLinkage  process in Excel (generate graphs, ...)
 Standard rules for LOD-scores




>3  significant linkage
2<LOD<3  suggestive linkage
-2<LOD<2  uninformative
<-2  significant absence of linkage
Interpreting LOD plots
5
5
4
4
3
3
2
2
1
1
0
0
0
0,1
0,2
0,3
0,4
0,5
-1
-1
-2
-2
-3
-3
-4
-4
-5
-5
5
5
4
4
3
3
2
2
1
1
0
0
0,1
0,2
0
0,1
0,2
0,3
0,4
0,5
0
0
0,1
0,2
0,3
0,4
0,5
-1
-1
-2
-2
-3
-3
-4
-4
-5
-5
0,3
0,4
0,5
Exercise – Part 3
 Generate one pedigree file containing all family
members of both families (use Global ID’s)
 Generate a genotype file for each of the tested markers
 Run SuperLink analysis with the right settings
 Evaluate results
Exercise – Part 3 > Results
Strengthen the evidence
 Analyze more family members
 Analyze more families
 Analyze flanking markers




Look for more informative markers that result in higher LOD-scores
A series of flanking markers allows for multipoint linkage analysis
A series of linked markers gives more confidence (subjective)
Flanking markers can also be used to fine-map the linkage interval
Determine the linkage interval
NL
?
L
NL
NL
L
?
...
L
?
NL
L
L
?
NL
candidate
region
Exercise 2: find the linkage interval
Post linkage
 Create a list of all the genes within the linkage interval
 NCBI map viewer
 UCSC (also for non-coding RNA’s)
 Evaluate known gene functions for relevance to the
investigated phenotype
 Sequence genes
 Start with those that seem the most relevant to the disorder
 Start with the coding regions
 Screen the entire region with capture sequencing
 Finding a mutation and proving its causality is the
ultimate proof