Pearson's marrow-pancreas syndrome. A multisystem mitochondrial disorder in infancy. J Clin Invest. Phenotypic spectrum associated with mutations of the mitochondrial polymerase gamma gene. Human mitochondrial DNA deletios associated with mutations in the gene encoding Twinkle, a phage T7 gene4-like protein localized in mitochondria. Wolf A, Cowen D. The cerebral atrophies and encephalomalacias of infancy and childhood.
POLG1 mutations associated with progressive encephalopathy in childhood. J Neuropathol Exp Neurol. Clayton DA. Replication of animal mitochondrial DNA. Spinazzola A, Zeviani M. Disorders of nuclear-mitochondrial intergenomic communication. Biosci Rep. Molecular diagnosis in mitochondrial complex I deficiency using exome sequencing. J Med Genet. Isolated complex I deficiency in children: clinical, biochemical and genetic aspects.
Hum Mutat. A common pattern of brain MRI imaging in mitochondrial diseases with complex I deficiency. Demonstration of a new pathogenic mutation in human complex I deficiency: a 5-bp duplication in the nuclear gene encoding the kD AQDQ subunit.
Smeitink J, van den Heuvel L. Human mitochondrial complex I in health and disease. The first nuclear-encoded complex I mutation in a patient with Leigh syndrome. Eur J Hum Genet. Shoubridge EA. Nuclear genetic defects of oxidative phosphorylation. Defective complex I assembly due to C20orf7 mutations as a new cause of Leigh syndrome. C6ORF66 is an assembly factor of mitochondrial complex I. Human CIA30 is involved in the early assembly of mitochondrial complex I and mutations in its gene cause disease.
Baysal BE. Hereditary paraganglioma targets diverse paraganglia. A mutant mitochondrial respiratory chain assembly protein causes complex III deficiency in patients with tubulopathy, encephalopathy and liver failure. Severe infantile encephalomyopathy caused by a mutation in COX6B1, a nucleus-encoded subunit of cytochrome c oxidase.
Mutations in COX10 result in a defect in mitochondrial heme A biosynthesis and account for multiple, early-onset clinical phenotypes associated with isolated COX deficiency.
Cytochrome c oxidase biogenesis in a patient with a mutation in COX10 gene. Mutations in COX7B cause microphthalmia with linear skin lesions, an unconventional mitochondrial disease. Reticulolinear aplasia cutis congenita of the face and neck: a distinctive cutaneous manifestation in several syndromes linked to Xp Br J Dermatol.
Exocrine pancreatic insufficiency, dyserythropoeitic anemia, and calvarial hyperostosis are caused by a mutation in the COX4I2 gene. SURF1, encoding a factor involved in the biogenesis of cytochrome c oxidase, is mutated in Leigh syndrome.
Loss-of-function mutations of SURF-1 are specifically associated with Leigh syndrome with cytochrome c oxidase deficiency. Mutations in C12orf62, a factor that couples COX I synthesis with cytochrome c oxidase assembly, cause fatal neonatal lactic acidosis.
A mutation in C2orf64 causes impaired cytochrome c oxidase assembly and mitochondrial cardiomyopathy. FASTKD2 nonsense mutation in an infantile mitochondrial encephalomyopathy associated with cytochrome c oxidase deficiency. Respiratory chain complex V deficiency due to a mutation in the assembly gene ATP Antenatal mitochondrial disease caused by mitochondrial ribosomal protein MRPS22 mutation.
Mitochondrial myopathy, sideroblastic anemia, and lactic acidosis: an autosomal recessive syndrome in Persian Jews caused by a mutation in the PUS1 gene. J Child Neurol. A human dynamin-related protein controls the distribution of mitochondria.
J Cell Biol. A lethal defect of mitochondrial and peroxisomal fission. A reduction of mitochondrial DNA molecules during embryogenesis explains the rapid segregation of genotypes. The implications of mitochondrial DNA copy number regulation during embryogenesis. Mitochondrial DNA mutations in human disease. Am J Med Genet. Performance comparison of exome DNA sequencing technologies. Nat Biotechnol. A mitochondrial protein compendium elucidates complex I disease biology. MitoP2: the mitochondrial proteome database—now including mouse data.
Whole-exome sequencing identifies homozygous AFG3L2 mutations in a spastic ataxia-neuropathy syndrome linked to mitochondrial m-AAA proteases. PLoS Genet. Mutation in PNPT1, which encodes a polyribonucleotide nucleotidyltransferase, impairs RNA import into mitochondria and causes respiratory-chain deficiency.
An RMND1 Mutation causes encephalopathy associated with multiple oxidative phosphorylation complex deficiencies and a mitochondrial translation defect. Identification of a mutation in LARS as a novel cause of infantile hepatopathy. Mol Genet Metab. Exome sequencing identifies MRPL3 mutation in mitochondrial cardiomyopathy. Exome sequencing identifies mitochondrial alanyl-tRNA synthetase mutations in infantile mitochondrial cardiomyopathy.
Titin mutation segregates with hereditary myopathy with early respiratory failure. The inheritance of mitochondrial DNA heteroplasmy: random drift, selection or both? Segregation of mtDNA throughout human embryofetal development: m. Treatment for mitochondrial disorders. Cochrane Database Syst Rev.
A randomized placebo-controlled trial of idebenone in Leber's hereditary optic neuropathy. Progress and prospects: gene therapy for mitochondrial DNA disease. Gene Ther. Assembly of functional proton-translocating ATPase complex in yeast mitochondria with cytoplasmically synthesized subunit 8, a polypeptide normally encoded within the organelle.
Allotopic mRNA localization to the mitochondrial surface rescues respiratory chain defects in fibroblasts harboring mitochondrial DNA mutations affecting complex I or v subunits.
Rejuvenation Res. Endurance training and detraining in mitochondrial myopathies due to single large-scale mtDNA deletions. Allogeneic hematopoietic SCT as treatment option for patients with mitochondrial neurogastrointestinal encephalomyopathy MNGIE : a consensus conference proposal for a standardized approach.
Bone Marrow Transplant. Mutations in the MPV17 gene are responsible for rapidly progressive liver failure in infancy.
PGD and heteroplasmic mitochondrial DNA point mutations: a systematic review estimating the chance of healthy offspring. Hum Reprod Update. Healthy baby girl born following pre-implantation Genetic diagnosis for mitochondrial DNA m.
Pronuclear transfer in human embryos to prevent transmission of mitochondrial DNA disease. Towards germline gene therapy of inherited mitochondrial diseases. Poulton J, Oakeshott P. Accompanying genetic variants from whole genome sequencing data which were generated as part of the Genomes Project Abecasis et al.
We used phase three data that was phased and imputed v5a. All RNA sequencing data derived from different projects were processed in the same way to ensure comparability across analyses. Following this, in order to minimize the likelihood of incorrectly placed reads particularly those associated with NUMT sequences , we used a stringent filtering pipeline, focusing only on reads that were properly paired and uniquely mapped.
To calculate transcript abundances, we used HTseq Anders et al. We also calculated principle components using the same data and removed outlier samples. We focused on mitochondrial encoded protein coding and ribosomal RNA genes only, since transfer RNAs showed lower sequencing coverage overall and were not expressed highly in all tissues and datasets.
For analysis of mitochondrial encoded gene expression variation across genes and datasets, for TwinsUK data we used only unrelated samples which involved picking one of each twin pair at random and combining these with unrelated samples. Genotyping data from different arrays and sequencing studies were processed separately. For TwinsUK data, only one twin from each twin pair was genotyped and thus processed, with data duplicated to represent the missing twin pair after quality control and filtering.
Genotyping quality control and calculation of genetic principle components for Twins data was thus performed only on unrelated samples. Problematic sites were removed and remaining SNPs were used for imputation in 2 MB intervals using impute2 Howie et al. Imputed data were then hard-called to produce genotypes at each site with a threshold of 0.
After processing, we calculated genetic principal components and removed outlier samples by visual inspection. For Geuvadis data we used whole genome sequencing variant calls from the Genomes project Abecasis et al. As such, these samples did not undergo phasing and imputation within our pipeline, but were filtered in the same way as genotyping data after this stage of the analysis. Expression QTL mapping was performed within each tissue and sequencing dataset. In each case, TPM values for thirteen mitochondrial encoded protein coding genes and two mitochondrial encoded ribosomal RNA genes were extracted before being log 10 transformed Supplementary file 4.
Mitochondrial encoded gene expression distributions were median normalized, before outlier values were removed per gene defined as three interquartile ranges above or below the upper and lower quartile respectively. For genotyping data, we restricted the data to only those samples that had corresponding mitochondrial encoded gene expression values for the given dataset and calculated genetic principle components on this reduced set in each case.
For twin data, we calculated the relatedness matrix of samples before conducting association analyses with GEMMA Zhou and Stephens, For TwinsUK data, the genotyping array was included as the batch covariate and sex was omitted as all samples were derived from females. After analysis, QQ plots were visually assessed and show no skew.
False discovery correction Benjamini-Hochberg was applied to raw p-values within each dataset by merging all genes 15 and genetic variants in each case, following the approach applied by the GTEx consortium GTEx Consortium et al.
In each case, we then collected the test statistic across all , permutations to generate a null distribution, and compared our observed test statistic against this to calculate an empirical P-value. For tissue types with multiple datasets Whole Blood and LCLs we performed permutations per dataset, combined these within a meta-analysis, and then derived the null distribution from the meta-analysis results.
In each case, we also then followed the approach outlined in Ongen et al. To do this, we performed random permutations across all nuclear genetic variants for the relevant mitochondria-encoded gene and tissue type, and then calculated the null distribution by selecting the largest test statistic per permutation across all nuclear genetic variants.
To calculate the overall family wise error rate, we repeated this again, this time selecting the largest test statistic across all nuclear genetic variants and all 15 mitochondria-encoded genes per permutation to generate the null distribution in the relevant tissue type. For the calculation of both family-wise error rates, we repeated the approach outlined in Ongen et al.
P values generated across all methods are shown in supplementary file 1. To test whether any 50 bp segments of mitochondrial genes also aligned to nuclear genes, we followed the approach defined in Saha and Battle Specifically, we took all 50 bp k-mers from each mitochondrial encoded gene and then aligned these sequences to the nuclear genome using bowtie v1.
For each nuclear genetic variant associated with a mitochondrial encoded gene, we then tested whether any of the 50 bp k-mers from the mitochondrial encoded gene aligned within a nuclear gene whose transcription start site fell within 1 MB of the corresponding nuclear genetic variant. For tissue types with multiple independent datasets, we defined discovery and replication datasets. Discovery datasets were chosen as the dataset with the largest starting sample size for each given tissue, with the replication dataset as the second largest.
For LCLs, where three independent datasets were available, we performed meta analysis combining data from the Twins and GTEx for the discovery phase, and then used Geuvadis data for replication. For all other tissues, only a single dataset was available, and so no replication analysis was performed.
We used the same approach when comparing association signals across tissues. To perform power calculations, we obtained the correlation coefficient r 2 between the genetic variant and the expression of the associated mitochondrial encoded gene in the relevant discovery dataset or largest dataset where the genetic variant is present, if multiple datasets are available for the tissue.
We then used a power calculator Purcell et al. Following this, we summed power values across all 61 associations. We also repeated all association analyses after using mitochondrial library size all reads mapping to the mitochondrial to calculate TPM for mitochondrial genes, rather than total library size. We tested this approach as a way to remove the effects of variable mitochondrial copy number and poly-cistronic transcription rate, however in all cases we obtained very similar results to those obtained using the method outlined above.
Additionally, we also repeated all analyses shown in Figure 1 using mitochondrial reads to normalize gene expression values; again we find very similar results. It has recently been shown that the post-mortem interval PMI appears to influence gene expression patterns in GTEx data Ferreira et al. As such, to test for an effect in our data, we repeated association analyses for significant associations discovered in GTEx data and including PMI as a covariate where PMI data were available.
In order to identify the potential causal nuclear gene associated with mitochondrial encoded gene expression, we identified genes associated with the peak eQTL variant in the following ways. First, if the peak variant was a missense mutation, we assumed that its mode of action was via functional changes in the gene it was located in. Second, for non-coding mutations, we tested whether non-coding peak variants fell in enhancer regions using chromatin state predictions obtained from cell types within the Roadmap Epigenetic project Kundaje et al.
Third, for non-coding peak variants, we tested for mediation via the expression of nuclear genes located near to the peak SNP. To do this, for each tissue we used the largest dataset available and restricted our analysis to unrelated samples for TwinsUK data, this involved picking one of each twin pair at random and combining these with unrelated samples. To prioritize potential causal genes within this framework, we first selected nuclear genes with a known role in mitochondrial processes any gene listed in the Mitocarta database Calvo et al.
To test whether associations between nuclear genetic variants and mitochondrial encoded gene expression that overlap GWAS signals are significant in individuals of European descent, we plotted the first two genetic principal components against those derived from genomes samples with known ancestry for any dataset that had associated RNA sequencing data from whole blood.
In order to validate the association between rs and expression levels of MTND4 in LCLs, we obtained ten LCL samples carrying the homozygous reference genotype and ten LCL samples carrying the homozygous non-reference genotype for rs from the Coriell Institute for Medical Research, matched between the two genotype groups for sex and ethnicity Supplementary file 5.
Outlier values were removed defined as three interquartile ranges above or below the upper and lower quartile respectively within each genotypic category, leaving 19 samples for analysis. This association was chosen for replication analysis since it is associated with mitochondrial encoded gene expression across multiple tissue types and is significantly associated with MTND4 in a dataset and tissue type for which we had access to the relevant biological material Geuvadis dataset, LCLs.
Anonymized processed mitochondrial encoded gene expression matrices are available in Supplementary file 2 and from the Gene Expression Omnibus under accession GSE In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses.
A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.
Thank you for submitting your article "Nuclear Genetic Regulation of the Human Mitochondrial Transcriptome" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by a guest Reviewing Editor and Mark McCarthy as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Dan Arking Reviewer 1. The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.
As you can see, we have listed a set of "essential revisions" that represent the consensus assessment of the editors and reviewers. The manuscript describes a study of the regulation of the mitochondrial transcriptome, in particular identifying nuclear genetic variants that are associated with expression of mitochondrial transcripts. They identify such associations for fourteen out of fifteen mitochondrial genes and a total of 64 trans-genome associations.
Replication between studies is observed, and they discuss the role of genetic variation affecting mitochondrial gene expression in complex disease. The reviewers appreciated the importance of the findings for understanding mitochondrial gene expression and found the manuscript generally clear.
However, to support the findings, some additional analyses are desired along with adding to the biological interpretation of the results and the discussion of some caveats. This could include an analysis of the variability in expression of those factors; a more systematic analysis of whether those factors are shuttled back to the nucleus or localized with the mitochondria; and analysis of the 45 QTLs that are not missense mutations.
For instance, are they in known enhancer regions that may have impacts on multiple mt-regulatory factors? Furthermore, because mitochondria function as the powerhouses of our cells, mitochondrial mutations often lead to more pronounced phenotypes in tissues that have high energy demands, such as brain, retinal, skeletal muscle, and cardiac muscle tissues. A number of clinical syndromes are currently believed to be associated with mitochondrial disease.
Possible examples include Pearson syndrome, Leigh syndrome, progressive external ophthalmoplegia, exercise-induced muscle pain , fatigue, and rhabdomyolysis. As previously mentioned, mitochondrial DNA in humans is always inherited from a person's mother Figure 4.
As a result, we share our mitochondrial DNA sequence with our mothers, brothers, sisters, maternal grandmothers, maternal aunts and uncles, and other maternal relatives. Due to the high mutation rates associated with mitochondrial DNA, significant variability exists in mitochondrial DNA sequences among unrelated individuals.
However, the mitochondrial DNA sequences of maternally related individuals, such as a grandmother and her grandson or granddaughter, are very similar and can be easily matched. Mitochondrial DNA sequence data has proved extremely useful in human rights cases, as it is a great a tool for establishing the identity of individuals who have been separated from their families.
This approach has been very successful for the following reasons Owens et al. One of the most prominent researchers to use mitochondrial DNA sequence data to tackle human rights issues is Dr. A particularly interesting example of Dr.
King's work occurred in Argentina. As a result of a military dictatorship that overthrew the existing Argentinean government in , thousands of citizens disappeared between and , including infants and children who were abducted along with their parents.
In addition, some children were born to women who were pregnant at the time of their kidnappings. After the military dictatorship was defeated, a new government commission predicted that at least 8, and possibly as many as 30, people had been kidnapped, including documented infants and children.
In , the grandmothers of these orphans formed the Associacion de Abuelas de Plaza de Mayo in an effort to identify their missing grandchildren, many of whom were illegally adopted by military families. In , King used mitochondrial DNA sequence data to reunite some of these Argentinean orphans with their grandmothers.
King collected blood samples from orphaned children and from women who had lost their children and grandchildren. Using mitochondrial DNA sequence data, she then matched more than 60 orphans with their biological families. In fact, as recently as , a young Argentinean man named Guillermo was finally reunited with his grandmother and sister.
Guillermo's parents were kidnapped by security forces in October ; Guillermo's mother, Patricia, was pregnant at the time of her kidnapping, and Guillermo was born one month later. Guillermo provided a blood sample to King's group, and his mitochondrial DNA sequence was a perfect match to that of one woman out of 2, in the database: Rosa, the mother of Patricia.
As an additional test, the researchers obtained a DNA sample from Mariana, the known daughter of Patricia, who was at a friend's house on the day her parents were kidnapped. As shown by this example, mitochondrial DNA sequences can be used to establish family ties with maternal relatives, even when both of a person's parents are missing.
Over the years, a probable role for mitochondria in both aging and cancer has emerged. As a byproduct of their role as powerhouses of our cells, mitochondria generate reactive oxygen species ROS. ROS production has been proposed to cause somatic mitochondrial mutations. This can lead to a cycle in which ROS generate mutations, which in turn lead to disregulation of respiration and accumulation of more mutations.
Indeed, ROS production contributes to tissue aging due to decreased metabolic function and energy production, increased cell death, and a decreased capacity to replicate the genome. In , a link between colorectal cancer and somatic mitochondrial mutations was established by Polyak and colleagues. These researchers cultured colorectal cancer cells taken from the tumors of 10 colorectal cancer patients. They then compared the mitochondrial DNA sequences of the tumor cell lines to the mitochondrial DNA sequences of cells from neighboring normal colon tissue from the same patient.
This side-by-side comparison was used to identify somatic mutations that had occurred in the mitochondrial DNA of the tumor cells. The researchers found that seven cell lines had acquired somatic mutations in their mitochondrial DNA sequences. Three of the cell lines had acquired a single mutation, and four had acquired between two and three mutations, for a total of twelve mutations. Eight of the mutations were in protein-encoding mitochondrial genes, and four of the mutations were in mitochondrial ribosomal RNA rRNA genes.
To confirm that the mitochondrial mutations had occurred in the tumors themselves and not during the culturing of the cells, the researchers next isolated DNA from the original tumor tissue and sequenced the mitochondrial DNA directly. Tumor tissue was only available for five out of the seven patients with mitochondrial mutations in their cultured cells. In all five cases, however, the same mutations were present in the primary tumor as in the cultured tumor cells.
Furthermore, the mitochondrial mutations were all homoplasmic in both the primary tumor and in the cultured cells. Based on these findings, Polyak and colleagues suggested that the somatic mitochondrial mutations might have provided a growth advantage to a single cell that subsequently proliferated more rapidly than the surrounding cells.
Furthermore, based on the homoplasmic nature of the mitochondrial DNA mutation, they suggested that the mutation might have provided a replicative advantage to the mutant mitochondrial genome. In the years that followed, a number of other studies also established associations between somatic mitochondrial mutations and various forms of cancer, including leukemias and solid tumors. However, a causative link between mitochondrial mutations and cancer has not yet been firmly established.
Clearly, the role of the mitochondrial genome must be considered with respect to human genetic disease. The heterogeneity of the mitochondrial genome presents many unmet challenges to researchers. However, emerging technologies are likely to aid the discovery of underlying genetic mechanisms linking these powerhouses to neurodegenerative disease, cancer, diabetes, and aging. Mitochondrial DNA plays important roles in other areas of genetics as well. For example, it has been used to address questions about how the widespread distribution of humans in the world today was established.
Because mitochondria are passed exclusively through the maternal lineage and there is little recombination in the mitochondrial genome, variation in the mitochondrial genome as well as in the Y chromosome in the case of paternal lineages has been used to delineate how and when humans migrated and occupied the world. Studying the mitochondrial genome in individuals from distinct geographic origins has made it possible to establish that the human populations of today are all derived from a small group of individuals that left Africa approximately , years ago Ingman et al.
Though small in size, the mitochondrial genome is responsible for ensuring that the powerhouses of our cells function properly. This circular genome is both more plentiful than its nuclear counterpart and more prone to mutation.
This is a particularly attractive interpretation, since these are the only proteins that are found in all characterized mitochondria Gray, Burger, and Lang There are other mitochondrial-encoded proteins that could conceivably provide this regulatory coupling. However, they all belong to the group of 64 proteins that are found in the mitochondrial genome of R. At this point, it is not possible to say with confidence why this small, variable core of proteins remains in some mitochondrial genomes but not in others.
William Martin, Reviewing Editor. Keywords: mitochondrial genome shrinkage population dynamic models for gene transfer. Address for correspondence and reprints: Otto G. E-mail: otto. Only the probability for the takeover by variant mN is shown. The solid curve uses the same parameter values as in figures 1—3. We thank Marguerite Picard for invaluable advice. We also thank her along with Siv Andersson for critical readings of the manuscript. Allen, J.
Control of gene expression by redox potential and the requirement for chloroplast and mitochondria genomes. Allen, C. Hakansson, and J. Redox conditions specify the proteins synthesized by isolated chloroplasts and mitochondria. Redox Rep. Andersson, S. An extreme codon preference strategy: codon reassignment.
Bioessays 14 — Reductive evolution of resident genomes. Trends Microbiol. Origins of mitochondria and hydrogenosomes. Zomorodipour, J. Andersson, T. Sicheritz-Ponten, U. Alsmark, R. Podowski, A. Eriksson, H. Winkler, and C. The genome sequence of Rickettsia prowazekii and the origin of mitochondria. Nature — Attardi, C. Animal mitochondrial DNA an extreme example of genetic economy.
Baker, K. Mitochondrial proteins essential for viability mediate protein transport into yeast mitochondria. Bergstrom, C. Germline bottlenecks and evolutionary maintenance of mitochondrial genomes.
Genetics — Bittner-Eddy, P. Monroy, and R. Expression of mitochondrial genes in the germinating conidea of Neurospora crassa. Bulmer, M. Coevolution of the genetic code and transfer RNA abundance. Campbell, C. Escape of mitochondrial DNA to the nucleus in yme1 yeast is mediated by vacuolar-dependent turnover of abnormal mitochondrial compartments.
Cell Sci. Covello, P. Silent mitochondrial and active nuclear genes for subunit 2 of cytochrome c oxidase cox2 in soybean: evidence for RNA-mediated gene transfer. EMBO J. Doolittle, R. Feng, S. Tsang, G. Cho, and E. Determining divergence times of the major kingdoms of living organisms with a protein clock. Science — Drake, J. A constant rate of spontaneous mutation in DNA-based microbes. USA 88 — Felsenstein, J. The evolutionary advantages of recombination. Genetics 78 — Feng, D.
Cho, and R. Determining divergence times 3ith a protein clock: update and reevaluation. USA 94 — Fox, T. Folley, J. Mulero, T. McMullin, P. Thorsness, L. Hedin, and M. Analysis of yeast mitochondrial genes.
Methods Enzymol. Galvis, M. Allen, and G. Moeller, P. Gardestroem, K. Glimelius, and E. Glaser, eds. Plant mitochondria: from gene to function. Backhuys, Leiden, The Netherlands. Gogarten, J. Olendzenski, and E. Dating the cenancestor of organisms. Gray, M. The endosymbiont hypothesis revisited. Burger, and B. Mitochondrial evolution. Hanekamp, T. Inactivation of YME2, an integral inner mitochondrial membrane protein, causes increased escape of DNA from mitochondria to the nucleus in Saccharomyces cerevisiae.
Cell 16 — Kadowski, K. Kubo, K. Ozawa, and A. Targeting presequence acquisition after mitochonrial gene transfer to the nucleus occurs by duplication of existing targeting signals. Keeling, P. Evidence that eukaryotic triosephosphate isomerase is of alpha-proteobacterial origin. Kimura, M. On the probability of fixation of mutant genes in populations. Genetics 47 — Kundig, C.
Henneke, and M.
0コメント