Construction of a linkage map of the zebra finch genome using SNP markers

Transcription

1 UPTEC X9 12 Examensarbete 3 hp Mars 29 Construction of a linkage map of the zebra finch genome using SNP markers Harriet Mellenius

2

3 Molecular Biotechnology Programme Uppsala University School of Engineering UPTEC X 9 12 Date of issue 29-3 Author Harriet Mellenius Title (English) Construction of a linkage map of the zebra finch genome using SNP markers Title (Swedish) Abstract The zebra finch (Taeniopygia guttata) is one of the most used model organisms for studies of behaviour and neurology, especially pertaining to the learning process. Nevertheless, genetic information about the zebra finch has until recently been scarce. The aim of this project was to produce a linkage map over the zebra finch genome. A pedigree of 1,351 birds was investigated, and 1,8 were successfully genotyped for 1,424 single nucleotide polymorphisms (SNPs). The linkage analysis resulted in a framework map consisting of 423 markers, covering 32 chromosomes and 1,34.2 cm. The results reveal that physical and genetic distances show a non-linear relationship in the zebra finch. The map will be used to trace the genetic background of the phenotypic traits recorded in this zebra finch pedigree. Keywords Evolutionary genetics, genetic linkage analysis, linkage map, zebra finch, recombination, single nucleotide polymorphism Supervisors Scientific reviewer Project name Niclas Backström Uppsala universitet Mikael Thollesson Uppsala universitet Sponsors Language ISSN English Supplementary bibliographical information Security Classification Pages 68 Biology Education Centre Biomedical Center Husargatan 3 Uppsala Box 592 S Uppsala Tel +46 () Fax +46 ()

4

5 Construction of a linkage map of the zebra finch genome using SNP markers Harriet Mellenius Sammanfattning En genetisk karta informerar om både hur en arts arvsmassa är arrangerad och var man ska söka efter de gener som påverkar en viss egenskap. Kartan placerar de genetiska markörer man undersöker i grupper motsvarande artens kromosomer och i ordning längs kromosomerna. Avstånden mellan markörerna i en grupp mäts i förekomsten av överkorsning mellan dem; hur ofta två nästan likadana kromosomer utbytt information. Av de två nästan likadana kromosomerna kommer en från mamman och en från pappan. Därför är en genetisk markör en bit av arvsmassan som ofta skiljer sig åt mellan individer så att den kan påvisa överkorsningarna. I denna studie användes variation av enstaka byggstenar i den genetiska koden som markörer. Kartan byggs genom att man i ett släktträd observerar hur ofta markörerna tillsammans förs vidare till nästa generation. Om två markörer nästan alltid ärvs ihop betyder det att de tillhör samma kromosom, och när de ibland skiljs åt har överkorsning skett mellan dem. I detta projekt användes ett släktträd med 1351 zebrafinkar vars arvsmassa undersöktes i 1424 punkter, och den karta som byggdes inkluderade 423 markörer. Hos dessa zebrafinkar hade också många beteenderelaterade egenskaper studerats, vars genetiska bakgrund kanske kommer att kunna utforskas med hjälp av kartan. Examensarbete 3hp Civilingenjörsprogrammet Molekylär bioteknik Uppsala universitet mars 29

6

7 ABBREVIATIONS 3 INTRODUCTION 4 GENETIC MAPS, A HISTORY 4 GENETIC MAPPING AND SEQUENCING IN AVIAN GENOMES 4 RECOMBINATION IS NECESSARY FOR GENETIC MAPPING 5 GENETIC MARKERS 6 MAPPING METHODS 7 THE ZEBRA FINCH 8 THE POPULATION 9 AIMS AND STRATEGY 9 METHODS 1 SNP IDENTIFICATION 1 GENOTYPING 1 DATA PROCESSING 11 MAPPING 11 GRAPHICS 13 PHYSICAL DISTANCES 13 RESULTS 14 MARKER GENOTYPING 14 LINKAGE GROUPS 14 THE FRAMEWORK MAP 15 COMPARISONS 2 THE GENETIC MAPS OF THE ZEBRA FINCH 2 THE CHICKEN GENOME 21 GENETICAL VS. PHYSICAL DISTANCES 25 DISCUSSION 28 CONCLUSIONS 3 ACKNOWLEDGEMENTS 3 REFERENCES 32 LIST OF APPENDICES 34 2

8 SNP cm RFLP QTL LOD BLAST Abbreviations Single Nucleotide Polymorphism centimorgan Restriction Fragment Length Polymorphism Quantitative Trait Loci Logarithm of Odds Basic Local Alignment Search Tool 3

9 Genetic maps, a history Introduction The aim of genetic mapping, unlike genetic sequencing, is not to determine every nucleotide in the genetic sequence, but to construct a general representation of a portion of the genome, consisting of the linear order of a set of selected markers. The map, genetic or physical, can give information about the karyotype and reveal the approximate position of the included markers. While a physical map will give the distance between markers in physical units, that is, the number of base pairs, the distances in a genetic map are based on the frequency of recombination events over a distance. Genetic mapping and sequencing in avian genomes The species that scientists have been most eager to genetically map are the ones that have been used extensively in previous research, called model organisms, in order to be able to compare previously obtained knowledge with new genetical data. The very first genetic maps of chromosomes were developed for the Drosophila melanogaster genome; one of the most commonly used model organisms (Sturtevant 1913). Other species of interest have been organisms of economical significance. Those are the reasons why the chicken (Gallus gallus) was the first, and at present the only, avian genome to be sequenced. Yet, the chicken is not representative for all birds and yields only limited insight to wild avian populations. The chicken belongs to the fowls, though the biggest of the avian orders containing nearly half of all avian species is Passeriformes, the passerines (Sibley and Ahlquist 199, Barker et al. 24), which have been the focus of many ecological studies. At this point, the only passerine birds genetically mapped for the whole genome are the zebra finch, for which a genetic map was published in May 28 (Stapley et al. 28), and the collared flycatcher in July 28 (Backström et al. 28). Partial maps of passerine genomes worth mentioning have been constructed for the great reed warbler (Hansson et al. 25, Åkesson et al. 27) and the house sparrow (Hale et al. 28). 4

10 Previous studies on chicken and passerine genomes have revealed an almost oneto-one homology between chicken and passerine chromosomes. The exceptions are chicken chromosomes 1 and 4, which are represented by two chromosomes each in the passerine genomes (Derjusheva et al. 24, Stapley et al. 28). However, even though the general chromosome organisation is preserved, there is evidence of intrachromosomal rearrangements (Stapley et al. 28, Backström et al. 28). Recombination is necessary for genetic mapping Recombination is a process that occurs uniquely in meiosis, after DNA replication. Homologous chromosomes, consisting of the two sister chromatids with a common centromere, pair up forming tetrads, where chiasmata can be formed along the chromosome arms. A chiasma is the complex where crossover, or recombination, occurs. Since the homologous chromosomes originate from the two parents, the recombined chromosome will be an assembly of genes from both parents. When the newly formed gamete results in offspring, each chromosome will be composed of genetic material from both grand-parents, either maternal or paternal. The unit of the genetic map is the centimorgan, cm. The distance of one cm between two linked markers implicates a 1% incidence that they are separated by recombination. The cm is related to the physical distance measured in base pairs in the sense that recombination is generally more likely to occur somewhere along a long stretch of the chromosome than over a shorter one, but they are not proportional as additional factors affect the rate of crossing-over. For instance, the recombination rate varies along the chromosome, with reduced rate of recombination close to the centromere. It has been suggested from studies on chicken that this tendency is weaker in birds than in other vertebrates (Jensen- Seaman et al. 24, Schmid et al. 25). An interesting effect of recombination is that where only one of the two chromosomes contains a variant, recombination can bring those variations 5

11 together onto the same chromosome, or, on the other hand, break up a collection of genetic variants, called a haplotype. When genes are linked, grouped together on the same chromosome, they are allowed to co-segregate through a pedigree. If the genes of a haplotype interact to give a combined effect, this effect may be observable in the pedigree, and may also be subject to selection. Thus, the mapping of genetic linkage can be of use for gene annotation, and regions where recombination seems to be underrepresented can be assumed to contain informative haplotypes. Genetic markers In order to distinguish the two chromosomes in a chromosome pair from each other, the genetic marker must be a site where variation is common, so that a large enough number of individuals in the investigated pedigree are heterozygous for that site. Only where it is possible to tell the alleles apart, recombination events can be detected. Throughout the history of genetic linkage analysis, different types of markers have been used. Among the first genetic markers to be used were the RFLPs, or restriction fragment length polymorphisms (Grodzicker et al. 1974). Restriction enzymes are used to cleave DNA strands at sites of short, specific sequences. Where there is variation among the restriction sites between the chromosomes, the DNA fragments resulting from restriction enzyme cleavage will be of distinctly different lengths. The major benefit with RFLPs is the ease with which new markers can be developed; the analysis is, on the other hand, slow and inconvenient. Sometimes, a nucleotide sequence can be repeated many times consecutively. The number of times the sequence is repeated varies among individuals, and they can thus be used as a genetic marker. If the sequence is very short, up to six base pairs, the repeat is called a microsatellite (Ellegren 24). Due to the presence of multiple alleles, microsatellite polymorphisms are so variable in number of repeats that they can even tell closely related individuals apart, and are therefore the genetic marker used in DNA profiling. The variability is clearly a 6

12 big advantage when microsatellites are used as genetic markers, but also makes comparisons between species troublesome. The genetic markers used in this study were single nucleotide polymorphism, or SNPs. A SNP is a substitution of a single nucleotide by mutation that makes a common variation in the population. Typically, the SNP nucleotide alternates between only two bases. The frequency of each of these in the population is called allele frequency. For a SNP to be suitable as a genetic marker, the allele frequency must be sufficiently high for it to be fairly common with biallelic individuals in the population. An advantage using SNPs as markers, that was a benefit to this study, is that the flanking sequences of the SNP is known, thus enabling searches for homologues in other sequenced species. Mapping methods Genetic linkage analysis is a statistical technique for building genetic maps from linkage data. Linkage data is typically obtained through genetic analysis of a pedigree where the segregation of markers can be observed, and recombination events traced. The collected data will tell how often a pair or a set of markers are inherited together. Allele combinations that are inherited together more often than expected are assumed to pertain to the same chromosome or in the same linkage group on a chromosome. Subsequently, the markers within each linkage group can be ordered. The order of the markers is based on how often recombination occurs between each pair of markers, and to be able to determine where a recombination event has chanced, the descent of the markers must be revealed. Only after deciding whether a chromosome is paternal or maternal, and the parental markers are genotyped, it is possible to detect recombination events. Recombination events are detected by the sudden change from grand-maternal to grand-paternal marker alleles along a chromosome, or vice versa. The fraction of observed recombinations out of all informative meioses estimates the recombination rate between a pair of markers, which is by definition the genetic 7

13 distance between them. However, there is always a risk that two or more recombination events occur between two adjacent markers. On these occasions, the total count of recombinations will be underestimated, as double recombinations re-change the alleles. This problem is dealt with by the recalculation of all recombination rates using a function that takes into account the chance of double crossovers between adjacent markers. Moreover, the probability of a recombination event is reduced close to other crossovers. This phenomenon is called crossover interference (Sturtevant 1915), and the mechanism behind it is not yet fully uncovered. The Kosambi mapping function take this as well into account, and this function is hence used to make observed recombination rates better represent the true recombination fractions along the chromosome (Kosambi 1944, Zhao & McPeek 1996). Using these data, the most likely marker order can be computed by addition of the calculated genetic inter-distances between the markers. This information can be ambiguous, so that there is seldom a single best order that includes all markers, but more often many plausible marker orders of similar likelihood. The concept of framework maps refers to that the final marker order has significantly higher probability compared to the alternatives. This map unlikely includes all markers as some can be placed at more than one locus with equal probability. The Zebra Finch The zebra finch (Taeniopygia guttata) may not be one of the most well-known model organisms, but is in fact the most used model organism for the study of behaviour and neurology, especially pertaining to the learning process (Jin & Clayton 1997, Bottjer et al. 1985). Every male bird learns to sing by imitating a tutor, much like human babies learn to speak by listening to grown-ups (Williams 24). An interesting aspect of zebra finch singing is the social context, such as the mating behaviour where the males sing in courtesy to the female. Consequently, the singing has been widely studied due to an interest in examining the genetic basis of learning. As a model organism, the zebra finch has the advantages of being easy to breed in captivity, and it is just as popular among pet owners as in scientific contexts. 8

14 A genetic map of the zebra finch genome will enable scientists to associate their behavioural and neurological data to the regions in the genome that affect the observed phenotype. Furthermore, the zebra finch genome is currently being sequenced ( which gives the opportunity to compare physical and genetic distance to exhibit variations in recombination rate over the chromosomes. The population The population of zebra finches used in this study was bred at the Department of Behavioural Ecology and Evolutionary Genetics at the Max Planck Institute for Ornithology at Seewiesen by Dr Wolfgang Forstmeier. The original ancestors were 63 males and 84 females, taken from the same population used in the zebra finch linkage map constructed by Stapley et al. (28). The ancestors, of unknown kinship, gave rise to a pedigree of 1,24 individuals, all interrelated. The pedigree consisted of 535 males, 524 females, and 145 offspring of unknown sex, the youngest separated from the ancestors by at most four generations. The birds were paired under controlled conditions. Maternity of offspring was determined by observation, and paternity, when uncertain, by DNA analysis of ten microsatellites. The entire pedigree contained 1,351 individuals. The most interesting feature of the pedigree is how well-studied it is. Numerous traits such as mass and tarsus length (relating to growth), beak colour and song rate (relating to attractiveness of males), and aggressiveness and responsiveness in mate choice situations (behavioural traits) have been monitored. A genetic map covering this specific pedigree will hopefully enable the localisation of approximate chromosome positions connected to these traits, which would be a significant scientific progress. Aims and Strategy The aim of the project was to produce a genetic map of the genome of the zebra finch. Such a map would be beneficial for, inter alia, gene anchoring in the zebra finch genome, in particular for genes involved in neurological functions, QTL 9

15 (quantitative trait loci) analysis and for comparative genomics of birds. This genetic map would be obtained by linkage analysis of recombination data from a pedigree of zebra finches, genotyped for some marker of preference. This project used SNPs that were previously identified. Methods The assignment of the degree project was the data analysis of the already genotyped SNPs, but to put it in its correct context, the previous steps of the mapping project need to be described. SNP identification The original SNP panel consisted of 1,92 SNPs in total; 617 SNPs from the SNP panel used by Stapley et al. (28) in the first zebra finch map, 187 identified from the sequencing of a number of pooled individuals from the pedigree in Sheffield, 917 from the genome sequencing of a single zebra finch of American origin, and 199 that had been identified in both of the latter reads. The newly identified SNPs were chosen from sequences that had a single known chicken homologue, as well as some of the Stapley SNPs. Hence 1,775 of the markers were homologues to loci with known position in the chicken genome. Genotyping The genotyping of the 1,92 SNPs was performed for 1,8 individuals from the pedigree, using the Golden Gate Assay (Fan et al. 23) from Illumina (San Diego) at the SNP Technology Platform in Uppsala, Uppsala University ( Prior to the genotyping, the DNA sequences flanking the SNPs were examined for nucleotide composition, other polymorphisms etc. to determine that they were appropriate as primers in the genotyping process. The quality of the genotype data and the SNPs as genetic markers was controlled by a set of quality control tests, including call rate, minor allele frequency, duplicate tests for reproducibility and detection of inheritance conflicts. The call rate of each SNP was calculated as the fraction of genotyping analyses that were successful, including duplicates, for all individuals. The minor allele frequency 1

16 shows whether the SNP is polymorphic in this population; if the minor allele frequency is zero, there is no variation to use for linkage analysis. An inheritance conflict is the deviation from Mendelian inheritance of the genotype data of parents and offspring. Inheritance conflicts in the data hence represent either genotyping errors or errors in the pedigree of the population. Data processing Prior to the linkage analysis, both the genotyping data and the pedigree had to be revised to assure linkage data quality. Among the markers, 18 out of 1,92 SNPs failed completely in all genotyping attempts. In addition, 38 SNPs had a minor allele frequency equal to zero and had to be discarded. Many inheritance conflicts were reported, but most of them erroneously, as no respect were taken to the fact that all markers on the Z chromosome were reported as homozygous. One individual however had so many inheritance errors that its pedigree position was considered wrong, and all its genotype data had to be removed. Fortunately, this had only minor effect on the analysis as this individual had neither siblings nor offspring. Eight SNPs had to be rejected due to excessive inheritance failures that indicated genotyping difficulties. Ultimately, 1,424 out of 1,92 markers persisted. The proportion of succeeded markers for the SNPs of different origin was 587/617 from the Sheffield panel (Stapley et al. 28), 144/187 from the sequencing on pooled individuals, 56/917 from the sequencing of a single American zebra finch and 187/199 of the combined. Obviously, the SNP identification in the reads from the single American individual did not conform with the pedigree in question, perhaps due to differences in allele frequencies between population as all other birds were related to the Sheffield population, or due to individual variations being mistaken for polymorphisms. Mapping The linkage analysis was performed with CRI-MAP v. 2.4 (Green et al. 199), but with additions and modifications by Xuelu Liu (Monsanto) in order to be able to manage the very complex pedigree in question. To start with, the TWOPOINT 11

17 program option was executed with LOD score > 3 (logarithm of odds, corresponding to a p-value of.1). TWOPOINT makes twopoint linkage analyses to establish linkage in between each pair of markers by calculating whether two markers co-segregate through the pedigree. Henceforth, markers were grouped by the command AUTOGROUP, an addition in the Monsanto version. AUTOGROUP places markers that display linkage to each other in the twopoint data output in linkage groups. Twopoint analysis for 1,424 markers can make up in 1,13,176 different pairs in theory. In practice, with a significant LOD score of at least 3 and linkage presumably only within linkage groups, there were only 5,714 pairs with significant linkage in the TWOPOINT analysis. However, with such an amount of data points, a probability of one false positive in a thousand would make a considerable amount of false linkages. A LOD score of 8 was consequently used for the AUTOGROUP command as a threshold for inclusion in a linkage group, producing only 1 expected false positive in each 1 8. This produced a set of linkage groups that corresponded well with the expected chromosomes. To include all significant linkages in the linkage groups, the AUTOGROUP option was executed for each linkage group separately using LOD score 3 within groups. In those executions, all markers suggested in first AUTOGROUP analysis were included, as well as the markers suggested with reference to the chicken chromosome organisation that had displayed significant linkage (of LOD score > 3) in the original TWOPOINT execution. The linkage groups contained at maximum 177 markers, which substantially reduced the risk of false positives. Finally, all markers that had not been included in the AUTOGROUP executions despite having significant linkage (but with unknown position in the chicken genome) were assigned to linkage groups by their twopoint linkage data output. Besides the LOD score, the AUTOGROUP command requires three additional parameters to be specified to sort the markers into linkage groups; the minimum number of informative meioses relative to the average number of informative 12

18 meioses, the maximum number of linkages to other groups, and minimum fraction of linkages for each marker that are to the assigned group. These parameters were set to, 2 and.3, respectively. Those constrains were all relaxed so that the LOD score was hence the only restraining parameter. When the linkage groups were established, the BUILD command that computes the most likely marker order with LOD score > 3 was run in each linkage group to produce a framework map. The BUILD command iteratively adds markers to the set of ordered markers, if the maximum likelihood of the order is improved by more than the threshold LOD score. Throughout the linkage analysis, distances were sex-averaged and calculated in Kosambi cm, with exception for TgZ (Taeniopygia guttata chromosome Z). When building TgZ, individuals of unknown sex were excluded, using only 1,26 individuals of which 935 were genotyped. All females appeared as homozygous for all loci due to females having only one Z chromosome, why one of the assumed alleles for each locus was dismissed in females. Since recombination on sex chromosomes only occurs in the homogametic sex, the distances calculated were sex-specific for males. Graphics Alignment of maps, as well as visualisation of maps, was performed using MapChart (Voorips 22). Physical distances For the comparison of genetic and physical distances, the flanking regions of the SNPs were used in BLAST searches in the zebra finch sequence assembly ( to obtain the physical positions of the SNPs. 13

19 Marker Genotyping Results The average success rate, i.e. sample call rate, among the residual markers was 95.4%. In this group, 18 out of 81,563 duplicates failed, giving a reproducibility of 99.98%. Linkage Groups The linkage group assignment resulted in 32 linkage groups, corresponding to chromosome 1-28 plus chromosome Z, with double linkage groups for Number of markers Number of markers in framework map Distance covered in Kosambi cm Linkage group Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg Tg TgZ Total Table 1. Number of markers by linkage group. LOD > 3. Linkage groups named according to chicken counterparts. chromosome 1, 24 and 25, in the chicken genome and named accordingly; where a chicken chromosome corresponded to more than one linkage group in the zebra finch linkage, those were simply numbered (Table 1). Using LOD score > 3, 1,44 of 1,424 markers displayed linkage to some linkage group. All markers included in the linkage groups are listed in Appendix A. For chicken chromosome 22, there was only a single marker with no linkage to any other group. It is included in the table of linkage groups anyhow, which thus contains 1,45 14

20 markers, owing to its presence in the former zebra finch map. Most of the markers were assigned to the expected chromosome, with only three exceptions; one marker located at chromosome 2 in chicken was found on chromosome 6 in zebra finch, and another, located at chromosome 1 in chicken, appeared on chromosome 14 in zebra finch, and lastly, a marker from chicken chromosome 4 was found in zebra finch chromosome 2. Linkage group sizes correspond roughly with chicken chromosome lengths. As expected, chicken chromosome 1 correspond to two big linkage groups. However, chromosome 4, which was previously divided into two linkage groups as well, is only represented by one linkage group in these results. This linkage group contains markers from both chromosome 4 linkage groups in the previous zebra finch map (Stapley et al. 28). The Framework Map The framework map provided by linkage analysis with LOD score > 3 contained 423 of the 1,44 linked markers (3.1%) and spanned 1,34.2 Kosambi cm. The average genetic distance between adjacent markers was 3.43 cm, with a standard deviation of 5.64 cm. Chromosome 22 was also included in the framework map (Table 1, Figure 1) for further comparisons to other maps. Table 1 provides a summary of the number of markers included in the framework map for each linkage group. To have only two markers ordered is obviously equal to have no markers ordered at all (as reading the linkage group from one end or the other makes no difference). Linkage groups with only two ordered markers have nevertheless been included in the framework map if the original linkage group contained only very few markers (eight or less), where comparisons to other maps are still informative. 15

21 16 TS1572 A16569 A12272 A13872 TS674 A18344 C145 A19836 F9931 A1578 TS17 TS1455 TS649 TS262 TS168 TS1188 TS16 C191 C198 A14883 A18836 TS832 TS444 C119 TS1451 A6985 TS816 C184 TS446 A1265 TS973 F1196 F11969 C188 C149 TS Tg1.1 TS1475 TS1538 C269 C243 TS82 TS453 TS959 C281 C384 TS85 A757 F27377 TS571 TS1289 C13 C134 C59 TS924 TS166 A4369 TS138 TS1453 A4243 C86 TS171 TS1291 TS898 A9491 TS815 TS958 TS Tg1.2 TS789 TS1458 TS744 TS43 TS1243 A3345 A41865 C399 TS257 TS745 C113 A42699 F2848 TS724 A7523 TS591 TS992 A1115 A13751 F1164 TS894 A13559 F9532 TS25 F13693 TS738 C25 TS Tg2 A1968 TS1441 A16738 F1496 TS558 F52 TS844 F5987 TS18 F6312 F2842 TS796 TS1211 TS169 TS365 A4172 TS1414 TS335 A485 A41675 A39343 A439 C369 A34284 A35359 TS422 A37893 A4132 A37111 A36468 A25769 A2377 A29198 A35582 A17389 A27742 A36 A28961 TS121 A3563 C381 A35839 A3742 TS Tg3 F15999 A29722 TS648 C364 TS81 TS129 A3652 TS122 A33982 TS1272 TS361 F1713 C272 C212 TS155 F14679 TS Tg4

22 17 TS368 TS289 TS391 TS1312 TS829 TS41 TS1292 TS459 C41 A3158 A4695 A43331 A5544 C7 TS166 A42516 C116 TS573 C456 F23839 TS111 TS1147 TS154 TS142 A22924 F1517 TS69 A2684 TS191 F Tg5 A955 TS1321 C12 TS21 A1528 TS822 A21 TS491 A25632 TS1629 A36826 TS23 F21486 TS431 A33993 TS987 TS786 A31359 TS812 A3151 F18753 A Tg6 TS178 A558 A4318 A42625 A5129 TS358 TS276 TS1224 A4829 TS486 A28355 A3113 C138 C24 A2745 F5452 TS166 TS1113 C56 TS Tg7 A27182 TS1494 A22212 A21974 A2355 A23631 A35525 A37197 TS264 A39248 A3856 A449 TS191 TS1382 C448 A39167 F2545 TS1153 A4627 C44 TS2 TS1464 TS Tg8 TS1419 TS115 F2377 C338 A3141 TS599 C284 TS93 A25561 A2776 A22253 A27226 A2112 C214 TS16 C4 C9 2 4 Tg9 TS39 TS95 F26842 A38288 A4455 A35438 C376 A2877 F21558 A3232 A22173 TS934 TS Tg1

23 18 TS99 A27945 TS797 C331 TS946 C343 F16314 TS1467 TS1431 TS971 TS Tg11 TS1476 A19389 A19735 A17518 A155 A1246 TS1422 TS156 A1462 F11646 F1615 TS911 TS1344 TS635 TS1454 TS Tg12 A771 C5 TS57 C228 TS87 TS1346 TS477 A2964 C215 C28 TS16 C Tg13 A32839 TS842 A294 TS389 TS256 A1451 TS Tg14 C414 F2514 TS585 TS16 TS333 A28843 TS341 C35 A Tg15 TS725 TS Tg16 TS388 TS732 A36774 A32346 TS115 TS369 TS54 TS35 A28575 TS542 TS Tg17 A34181 TS1125 F268 F22993 F Tg18 TS864 F22829 A28744 A18395 A32569 TS768 TS235 TS997 TS1253 A2711 C222 TS1158 F18227 F16699 TS561 TS59 TS Tg19 A22 TS8 A22881 C241 F23558 TS159 TS1119 TS92 A31598 TS1325 TS1569 TS912 TS1324 A Tg2

24 Tg21 Tg22 Tg23 Tg24.1 Tg24.2 Tg25.1 F15517 TS787 TS627 A1734 TS1315 TS948 TS675 TS1263 A2955 TS655 A2187 TS662 2 F TS143 TS1256 TS149 A23214 C7 Tg25.2 Tg26 Tg27 Tg28 TgZ C218 TS1376 TS1117 C264 TS1385 TS758 TS149 A12246 A24644 A825 TS142 TS A34858 A384 A A3446 A291 TS126 A TS1371 Figure 1. Linkage map over the zebra finch genome. The markers are ordered by LOD score > 3. Distances are measured in Kosambi cm. 19

25 The framework map (Figure 1) shows an obvious condensation of markers somewhere in the middle of the chromosome for many linkage groups. This condensation is assumed to represent chromosome areas close to the centromere, where theory states that fewer recombination events occur and genetic distances are thereby shorter. Comparisons The genetic maps of the zebra finch The genetic map of the zebra finch developed in this project had 587 markers in common with the previous genetic map by Stapley et al. (28). Out of these, 191 are included in the framework map. An alignment of the two zebra finch maps exposes the good coherence between them, as is exemplified with linkage groups Tg1.1-2, Tg2 and Tg3 in Figure 2. Tg1.1 Tgu1B Tg1.2 Tgu1A Tg2 Tgu2 Tg3 Tgu3 Tgu1 Tgu3A Tgu1C Figure 2. Linkage groups Tg1.1-2, Tg2, and Tg3; a comparison between the framework map (to the left) and the map by Stapley et al. (28) (to the right). All markers are included and common markers are connected. Tg1.1 and Tg3 have more than one homologous linkage group in the other map. 2

26 The chromosome assignment of the markers corresponds perfectly between the two maps, and marker order is almost identical. Due to the higher power of the linkage analysis of the Uppsala map, linkage was established to 7 linkage groups (in Tg1.1, Tg3, Tg16, Tg2, Tg25 and Tg26) that were out-groups in the Sheffield map (see Appendix B). The homologues to chicken chromosome 4, which was represented by two linkage groups in the Sheffield map but only by one in this study, would have been interesting to align, but unfortunately, all markers in Tg4 in the framework map corresponded to the linkage group Tgu4A in the Sheffield map. The whole alignment between the two zebra finch maps can be found in Appendix B. The chicken genome As the zebra finch SNPs were chosen in genes that were known to have chicken orthologues, almost all of the markers in the framework map could be linked to its chicken homologue. The linkage groups were, as mentioned above, almost perfectly preserved between chicken and zebra finch with only a few exceptions. There was, however, some disparity in gene order as is evinced in figure 3 that aligns the framework map with a physical map over the chicken genome and connects homologues. 21

27 Ggal1 Tg1.1 Ggal2 Tg2 Ggal3 Tg3 Ggal4 Tg4 Tg1.2 22

28 Ggal5 Tg5 Ggal6 Tg6 Ggal7 Tg7 Ggal8 Tg8 Ggal9 Tg9 Ggal1 Tg1 Ggal11 Tg11 Ggal12 Tg12 Ggal13 Tg13 Ggal14 Tg14 Ggal15 Tg15 Ggal16 Tg16 Ggal17 Tg17 23

29 Ggal18 Tg18 Ggal19 Tg19 Ggal2 Tg2 Ggal21 Tg21 Ggal22 Tg22 Ggal23 Tg23 Ggal24 Tg24.1 Ggal25 Tg25.1 Tg24.2 Tg25.2 Ggal26 Tg26 Ggal27 Tg27 Ggal28 Tg28 GgalZ TgZ Figure 3. Alignment of zebra finch framework map (right) and the chicken homologues (left). The chicken markers are interspersed by physical distances. Chicken chromosome 1, 24 and 25 are represented by two linkage groups each in the zebra finch framework map. Homologue sequences are connected. 24

30 Genetical vs. physical distances In the BLAST search for physical positions of the SNPs, 1,48 out of the 1,424 markers (98.9%) were found. However, some of the sequences yielded duplicate hits. Hits on other chromosomes than the assigned were dismissed. Duplicates very close to each other, within ~2 kbp (much closer than any two markers), were considered as one. Duplications that were further apart were not included in the comparisons. The genetic distance was plotted against the physical distance for all linkage groups containing more than 15 markers in the framework map; linkage groups Tg1.1-2, Tg2-9, Tg12 and Tg19. Figure 4 displays the plots for linkage groups Tg1.1-2, Tg2, Tg4, Tg12 and Tg19. The rest of the plots can be found in Appendix C. Tg Mbp cm 25

31 Tg Mbp cm Tg2 Mbp cm 26

32 Tg Mbp cm Tg Mbp cm 27

33 Tg Mbp cm Figure 4. Physical distance plotted against genetic distance for linkage groups Tg1.1-2, Tg2, Tg4, Tg12 and Tg19. Some linkage groups, such as Tg1.1-2 and Tg2-5, showed an apparent s-shaped curve as expected if recombination rate is negatively correlated with the distance to the centromere. For others, such as Tg19, the relation was a very evident straight line, indicating no such correlation. Some of the plots, like Tg12, were difficult to interpret, and, naturally, this concerned the plots with few data points in particular. Markers that appear outside the general curve or line are the ones for which the genetical map and the zebra finch sequence are not in agreement. None of the displayed plots contained any obvious out-groups, but in most plots, there were a few data points that call for closer investigation in the future, such as the two trios of markers that seem to have changed places at around 6 cm in Tg2. Discussion The result of the linkage analysis was a linkage map containing 3% of the markers, spanning 31 linkage groups. This map was compared with the previous 28

34 zebra finch linkage map (Stapley et al. 28) and with physical positions of the markers from the sequencing. The marker order was in general consentient. The framework map was also aligned with a physical map of the chicken genome. The most significant disparity between the previous zebra finch map and the linkage groups from this study regarded the equivalent of chicken chromosome 4, which had been divided into two linkage groups. This was not only proposed for zebra finch by Stapley et al. (28), but also by Derjusheva et al. (24) for domestic pigeon and the two passerines chaffinch and redwing, by Backström et al. (28) for the collared flycatcher, and others (Guttenbach et al. 23). The fact that chicken chromosome 4 was represented by only one linkage group in this study does not rule out the possibility that it is also homologous to another microchromosome. The only contradiction to previous results is the markers that were included in both this study and the map by Stapley et al. (28), which are believed to belong to the same linkage group when a larger dataset provides more power to the calculations. As mentioned above, the chicken does not belong to the group of passerines (species of the order Passeriformes), but to the fowls. This means that the chicken and the zebra finch are separated by at least ~1 million years of evolution (Sibley and Ahlquist 199, Barker et al. 24). The genetic map of the zebra finch yields new insight in how the two lineages have evolved since the split between them. Further studies might reveal how these changes are associated with the phenotypic differences between the fowls and the passerines. It has been suggested that the chromosome organisation of markers in avian genomes are preserved, but not the synteny within chromosomes (Schmid et al. 25). The first genetic map of the zebra finch was in support of this statement (Stapley et al. 28). The alignment of the Uppsala zebra finch framework map and the chicken genome provides further support, as was expected from the consensus of the two zebra finch maps. 29

35 In some chromosomes, such as chromosome 2, synteny seems to be remarkably well-preserved despite the 1 million years that separates the two species. In other, such as chromosome 15, there are rearrangements; however, the rearrangements seem to be explainable by only a few major inversions, in this example three. The hypothesis that the in other vertebrates observed correlation between recombination rate and distance to the centromere (Jensen-Seaman et al. 24) is not so strong in avian genomes (Schmid et al. 25), based on the chicken genome, was definitely contradicted by the physical-genetic distance plots of zebra finch linkage groups Tg1.1-2 and Tg2-5. However, the tendency was not as evident for all linkage groups, and not supported at all by Tg19. A possible explanation is that microchromosomes are too small to demonstrate a pronounced effect. It can nevertheless be concluded that the negative correlation between recombination rate and distance to the centromere is applicable to at least some avian genomes, if not chicken. Conclusions The new framework map over the zebra finch genome has already provided new insights in the passerine genome. The alignment against the chicken genome reveals chromosomal rearrangements that have occurred since the divergence of the orders of the fowls and the passerines. The physical-genetic distance plots in the zebra finch linkage groups shed light on the correlation between physical and genetic distances over the chromosome. The natural progression of the project is to construct a best-order map including all markers that were sorted into the linkage groups. These results will hopefully be of use in exploring the genetic background to the phenotypic traits recorded in this zebra finch pedigree. Acknowledgements First, I want to thank my supervisor Niclas Backström for guidance and support, and Professor Hans Ellegren, who initialised the project and provided valuable 3

36 advice. I thank Mikael Thollesson for accepting the task of the scientific reviewer, and Maryam Montazerolghaem and Martin Dahlö for kindly accepting to be my opponents. The whole research group of Professor Ellegren at the Department for Evolutionary Biology was a support in creating a good scientific work environment with many sensible ideas and opinions. I want to thank Axel Künstner in particular for providing invaluable technical assistance. Mathieu Authier s help with the software was also appreciated. Dr Wolfgang Forstmeier with colleagues at the Department of Behavioural Ecology and Evolutionary Genetics at the Max Planck Institute for Ornithology at Seewiesen are gratefully acknowledged for the breeding and the sampling of the pedigree. I thank Tomas Axelsson and his colleagues at the SNP Technology Platform at Uppsala University for the thorough SNP genotyping. Finally, Rickard Hedman is thanked for all his considerate help during the writing of this report. 31

37 References Åkesson, M., B. Hansson, D. Hasselquist and S. Bensch, 27. Linkage mapping of AFLP markers in a wild population of great reed warblers: importance of heterozygosity and number of genotyped individuals. Mol. Ecol. 16: Backström, N., N. Karaiskou, E. H. Leder, L. Gustafsson, C. R. Primmer, A. Qvarnström and H. Ellegren, 28. A Gene-Based Genetic Linkage Map of the Collared Flycatcher (Ficedula albicollis) Reveals Extensive Synteny and Gene- Order Conservation During 1 Million Years of Avian Evolution. Genetics 179: Barker, F. K., A. Cibois, P. Schikler, J. Feinstein and J. Cracraft, 24. Phylogeny and diversification of the largest avian radiation. Proc. Natl. Acad. Sci. USA 11: Bottjer, S. W., S. L. Glaessner and A. P. Arnold, Ontogeny of brain nuclei controlling song learning and behavior in zebra finches. J. Neurosci. 5: Derjusheva, S., A. Kurganova, F. Habermann and E. Gaginskaya, 24. High chromosome conservation detected by comparative chromosome painting in chicken, pigeon and passerine birds. Chromosome Res. 12: Ellegren, H., 24. Microsatellites: simple sequences with complex evolution. Nat. Rev. Genet. 5: Fan, J. B., A. Oliphant, R. Shen, B. G. Kermani, F. Garcia et al., 23. Highly parallel SNP genotyping. Cold Spring Harbor Symp. Quant. Biol. 68: Green, P., K. Falls and S. Crook, 199. Documentation for CRIMAP, Version 2.4. Washington University School of Medicine, St. Louis. Grodzicker T., J. Williams, P. Sharp and J. Sambrook, Physical mapping of temperaturesensitive mutations of adenoviruses. Cold Spring Harbor Symp. Quant. Biol. 39: Guttenbach M., I. Nanda, W. Feichtinger, J. S. Masabanda, D. K. Griffin and M. Schmid, 23. Comparative chromosome painting of chicken autosomal paints 1 9 in nine different bird species. Cytogenet. Genome Res. 13: Hale M, H. Jensen, T. Birkhead, T. Burke and J. Slate, 28. A comparison of synteny and gene order on the homologue of chicken chromosome 7 between two passerine species and between passerines and chicken. Cytogenet. Genome Res. 121:

38 Hansson, B., M. Åkesson, J. Slate and J.M. Permberton, 25. Linkage mapping reveals sex-dimorphic map distances in passerine bird. Proc. R. Soc. Lond. B. 272: Jensen-Seaman M. I., T. S. Furey, B. A. Payseur, Y. Lu, K. M. Roskin, C. F. Chen, M. A. Thomas, D. Haussler and H. J. Jacob, 24. Comparative recombination rates in the rat, mouse, and human genomes. Genome Res. 14: Jin, H. and D. F. Clayton, Localized changes in immediate-early gene regulation during sensory and motor learning in zebra finches, Neuron 19: Kosambi D. D., The estimation of the map distance from recombination values. Ann. Eugen. 12: Schmid, M., I. Nanda, H. Hoehn, M. Schartl, T. Haaf et al., 25. Second report on chicken genes and chromosomes 25. Cytogenet. Genome Res. 19: Sibley, C. and J. Ahlquist, 199. Phylogeny and Classification of Birds: A Study in Molecular Evolution. Yale University Press, New Haven, CT. Stapley, J., T. R. Birkhead, T. Burke and J. Slate, 28. A Linkage Map of the Zebra Finch Taeniopygia guttata Provides New Insights Into Avian Genome Evolution. Genetics 179: Sturtevant, A. H., The linear arrangement of six sexlinked factors in Drosophila, as shown by their mode of association. J. Exp. Zool. 14: Sturtevant, A. H., The behavior of the chromosomes as studied through linkage. Z. Indukt. Abstammungs. Vererbungsl. 13: Voorrips, R. E., 22. MapChart: software for the graphical presentation of linkage maps and QTLs. J. Hered. 93: Williams, H., 24. Birdsong and singing behavior, Ann. N. Y. Acad. Sci. 116:1 3. Zhao H. and M. S. McPeek, On genetic map function. Genetics 142:

39 List of appendices Appendix A. List of all markers included in the linkage groups (Appendices, p.1) Appendix B. Complete comparison with Sheffield map (Appendices, p.28) Appendix C. The remaining physical-genetic distance plots (Appendices, p.31) 34

40 Appendices Appendix A. List of all markers included in the linkage groups SNP Chicken chromosome Zebra finch linkage group In framework map? A371 chr1 Tg1.1 A6985 chr1 Tg1.1 yes A6992 chr1 Tg1.1 A7879 chr1 Tg1.1 A12263 chr1 Tg1.1 A12264 chr1 Tg1.1 A12272 chr1 Tg1.1 yes A12274 chr1 Tg1.1 A12281 chr1 Tg1.1 A1239 chr1 Tg1.1 A1265 chr1 Tg1.1 yes A1299 chr1 Tg1.1 A132 chr1 Tg1.1 A1336 chr1 Tg1.1 A1363 chr1 Tg1.1 A13629 chr1 Tg1.1 A13669 chr1 Tg1.1 A13872 chr1 Tg1.1 yes A1492 chr1 Tg1.1 A1454 chr1 Tg1.1 A14657 chr1 Tg1.1 A14883 chr1 Tg1.1 yes A14891 chr1 Tg1.1 A1491 chr1 Tg1.1 A14941 chr1 Tg1.1 A1578 chr1 Tg1.1 yes A15165 chr1 Tg1.1 A15394 chr1 Tg1.1 A15865 chr1 Tg1.1 A1615 chr1 Tg1.1 A16569 chr1 Tg1.1 yes A16886 chr1 Tg1.1 A16982 chr1 Tg1.1 A1739 chr1 Tg1.1 A1759 chr1 Tg1.1 A17722 chr1 Tg1.1 A18161 chr1 Tg1.1 A18339 chr1 Tg1.1 A18344 chr1 Tg1.1 yes A18719 chr1 Tg1.1 A18836 chr1 Tg1.1 yes A19234 chr1 Tg1.1 A19836 chr1 Tg1.1 yes A2288 chr1 Tg1.1 A43644 chr1 Tg1.1 C35 chr1 Tg1.1 C119 chr1 Tg1.1 yes C136 chr1 Tg1.1 35

41 C139 chr1 Tg1.1 C142 chr1 Tg1.1 C143 chr1 Tg1.1 C145 chr1 Tg1.1 yes C147 chr1 Tg1.1 C149 chr1 Tg1.1 yes C152 chr1 Tg1.1 C154 chr1 Tg1.1 C155 chr1 Tg1.1 C158 chr1 Tg1.1 C171 chr1 Tg1.1 C172 chr1 Tg1.1 C174 chr1 Tg1.1 C178 chr1 Tg1.1 C18 chr1 Tg1.1 C181 chr1 Tg1.1 C183 chr1 Tg1.1 C184 chr1 Tg1.1 yes C188 chr1 Tg1.1 yes C191 chr1 Tg1.1 yes C198 chr1 Tg1.1 yes C199 chr1 Tg1.1 C22 chr1 Tg1.1 C27 chr1 Tg1.1 F4399 chr1 Tg1.1 F673 chr1 Tg1.1 F8918 chr1 Tg1.1 F9243 chr1 Tg1.1 F9317 chr1 Tg1.1 F9931 chr1 Tg1.1 yes F1196 chr1 Tg1.1 yes F11425 chr1 Tg1.1 F11429 chr1 Tg1.1 F11969 chr1 Tg1.1 yes F11984 chr1 Tg1.1 F12228 chr1 Tg1.1 F12778 chr1 Tg1.1 F13467 chr1 Tg1.1 TS16 chr1 Tg1.1 yes TS17 chr1 Tg1.1 yes TS92 chr1 Tg1.1 TS154 chr1 Tg1.1 TS168 chr1 Tg1.1 yes TS196 chr1 Tg1.1 TS197 chr1 Tg1.1 TS22 chr1 Tg1.1 TS254 chr1 Tg1.1 TS259 chr1 Tg1.1 TS262 chr1 Tg1.1 yes TS384 chr1 Tg1.1 TS4 chr1 Tg1.1 TS48 chr1 Tg1.1 TS429 UN Tg1.1 TS433 chr1 Tg1.1 36

42 TS444 chr1 Tg1.1 yes TS446 chr1 Tg1.1 yes TS454 UN Tg1.1 TS455 UN Tg1.1 TS478 chr1 Tg1.1 TS487 chr1 Tg1.1 TS524 chr1 Tg1.1 TS556 chr1 Tg1.1 TS649 chr1 Tg1.1 yes TS665 chr1 Tg1.1 TS674 chr1 Tg1.1 yes TS681 chr1 Tg1.1 TS757 chr1 Tg1.1 TS759 chr1 Tg1.1 TS761 chr1 Tg1.1 TS816 chr1 Tg1.1 yes TS832 chr1 Tg1.1 yes TS836 chr1 Tg1.1 TS862 chr1 Tg1.1 TS915 chr1 Tg1.1 TS939 chr1 Tg1.1 yes TS961 chr1 Tg1.1 TS973 UN Tg1.1 yes TS974 UN Tg1.1 TS12 chr1 Tg1.1 TS18 chr1 Tg1.1 TS13 chr1 Tg1.1 TS113 chr1 Tg1.1 TS118 chr1 Tg1.1 TS1115 chr1 Tg1.1 TS1126 chr1 Tg1.1 TS1188 chr1 Tg1.1 yes TS1225 chr1 Tg1.1 TS1236 chr1 Tg1.1 TS1246 chr1 Tg1.1 TS131 chr1 Tg1.1 TS131 chr1 Tg1.1 TS1366 UN Tg1.1 TS1433 UN Tg1.1 TS145 UN Tg1.1 TS1451 UN Tg1.1 yes TS1455 UN Tg1.1 yes TS1479 UN Tg1.1 TS1515 UN Tg1.1 TS1536 UN Tg1.1 TS1572 UN Tg1.1 yes TS1582 UN Tg1.1 TS1678 chr1 Tg1.1 A4369 chr1 Tg1.2 yes A4763 chr1 Tg1.2 A492 chr1 Tg1.2 A5727 chr1 Tg1.2 A757 chr1 Tg1.2 yes A7733 chr1 Tg1.2 37

43 A8195 chr1 Tg1.2 A9491 chr1 Tg1.2 yes A9775 chr1 Tg1.2 A1145 chr1 Tg1.2 A11361 chr1 Tg1.2 A22128 chr1 Tg1.2 A27553 chr1 Tg1.2 A27798 chr1 Tg1.2 A28548 chr1 Tg1.2 A2872 chr1 Tg1.2 A29647 chr1 Tg1.2 A31676 chr1 Tg1.2 A32293 chr1 Tg1.2 A3265 chr1 Tg1.2 A35146 chr1 Tg1.2 A37946 chr1 Tg1.2 A39183 chr1 Tg1.2 A39858 chr1 Tg1.2 A41768 chr1 Tg1.2 A42364 chr1 Tg1.2 A42367 chr1 Tg1.2 A42413 chr1 Tg1.2 A4243 chr1 Tg1.2 yes A43247 chr1 Tg1.2 C37 chr1 Tg1.2 C38 chr1 Tg1.2 C48 chr1 Tg1.2 C59 chr1 Tg1.2 yes C72 chr1 Tg1.2 C83 chr1 Tg1.2 C86 chr1 Tg1.2 yes C13 chr1 Tg1.2 yes C134 chr1 Tg1.2 yes C243 chr1 Tg1.2 yes C269 chr1 Tg1.2 yes C278 chr1 Tg1.2 C279 chr1 Tg1.2 C281 chr1 Tg1.2 yes C384 chr1 Tg1.2 yes C385 chr1 Tg1.2 C393 chr1 Tg1.2 C394 chr1 Tg1.2 C436 chr1 Tg1.2 C449 chr1 Tg1.2 F2396 chr1 Tg1.2 F428 chr1 Tg1.2 F495 chr1 Tg1.2 F5777 chr1 Tg1.2 F6395 chr1 Tg1.2 F7555 chr1 Tg1.2 F19832 chr1 Tg1.2 F281 chr1 Tg1.2 F27252 chr1 Tg1.2 F27377 chr1 Tg1.2 yes 38

44 TS91 chr1 Tg1.2 TS171 chr1 Tg1.2 yes TS199 chr1 Tg1.2 TS27 chr1 Tg1.2 TS223 chr1 Tg1.2 TS227 chr1 Tg1.2 TS239 chr1 Tg1.2 TS261 chr1 Tg1.2 TS326 chr1 Tg1.2 TS364 chr1 Tg1.2 TS373 chr1 Tg1.2 TS47 chr1 Tg1.2 TS418 chr1 Tg1.2 TS453 chr1 Tg1.2 yes TS468 chr1 Tg1.2 TS472 chr1 Tg1.2 TS529 chr1 Tg1.2 TS553 chr1 Tg1.2 TS571 chr1 Tg1.2 yes TS576 chr1 Tg1.2 TS683 chr1 Tg1.2 TS692 chr1 Tg1.2 TS766 chr1 Tg1.2 TS82 chr1 Tg1.2 yes TS85 chr1 Tg1.2 yes TS815 chr1 Tg1.2 yes TS849 chr1 Tg1.2 TS878 chr1 Tg1.2 TS898 chr1 Tg1.2 yes TS924 chr1 Tg1.2 yes TS958 chr1 Tg1.2 yes TS959 chr1 Tg1.2 yes TS982 chr1 Tg1.2 TS186 chr1 Tg1.2 TS1136 chr1 Tg1.2 TS1181 chr1 Tg1.2 TS1265 chr1 Tg1.2 TS1286 chr1 Tg1.2 TS1289 chr1 Tg1.2 yes TS1291 chr1 Tg1.2 yes TS138 chr1 Tg1.2 yes TS1333 UN Tg1.2 TS1386 UN Tg1.2 TS1393 UN Tg1.2 yes TS1418 UN Tg1.2 TS1453 UN Tg1.2 yes TS1475 UN Tg1.2 yes TS1538 UN Tg1.2 yes TS166 chr1 Tg1.2 yes A324 chr2 Tg2 A3277 chr2 Tg2 A3335 chr2 Tg2 A379 chr2 Tg2 A3885 chr2 Tg2 39

45 A5152 chr2 Tg2 A5449 chr2 Tg2 A5759 chr2 Tg2 A6221 chr2 Tg2 A6365 chr2 Tg2 A6418 chr2 Tg2 A7256 chr2 Tg2 A7391 chr2 Tg2 A7523 chr2 Tg2 yes A7672 chr2 Tg2 A7779 chr2 Tg2 A88 chr2 Tg2 A8187 chr2 Tg2 A8758 chr2 Tg2 A8971 chr2 Tg2 A8988 chr2 Tg2 A9151 chr2 Tg2 A9979 chr2 Tg2 A1115 chr2 Tg2 yes A1348 chr2 Tg2 A1361 chr2 Tg2 A11299 chr2 Tg2 A11463 chr2 Tg2 A11861 chr2 Tg2 A11985 chr2 Tg2 A12481 chr2 Tg2 A13559 chr2 Tg2 yes A13751 chr2 Tg2 yes A1472 chr2 Tg2 A1598 chr2 Tg2 A1531 chr2 Tg2 A15936 chr2 Tg2 A15969 chr2 Tg2 A16924 chr2 Tg2 A17365 chr2 Tg2 A17555 chr2 Tg2 A1875 chr2 Tg2 A18942 chr2 Tg2 A1961 chr2 Tg2 A19258 chr2 Tg2 A19271 chr2 Tg2 A19295 chr2 Tg2 A19759 chr2 Tg2 A19954 chr2 Tg2 A2726 chr2 Tg2 A25817 chr4 Tg2 A29441 chr2 Tg2 A29465 chr2 Tg2 A3345 chr2 Tg2 yes A36138 chr2 Tg2 A36166 chr2 Tg2 A37629 chr2 Tg2 A37782 chr2 Tg2 A39297 chr2 Tg2 4