# Physical Mapping: How Genome Maps Are Built Physical mapping is the process of determining the actual order and spacing of DNA landmarks along a chromosome, expressed in base pairs or in the order of overlapping cloned fragments rather than in recombination units. A physical map answers the question "how far apart are these two sequences in the [DNA molecule](/blog/guides/dna-molecule) itself," while a genetic map answers "how often do these two sequences get separated by recombination." That distinction matters because the two maps measure different things and disagree in predictable ways. A genetic map compresses and expands regions depending on how often crossing over happens there. A physical map does not care about crossing over at all. When you want to clone a gene, validate a genome assembly, or walk from a marker to a causal variant, you need the physical map. When you want to know whether two traits assort independently, you need the genetic map. Most modern genome projects build both and integrate them. This guide covers how physical maps are constructed, the main workflows (restriction-site mapping, sequence-tagged site anchoring, clone fingerprinting, and contig assembly), the pitfalls that break assemblies, and how to read the numbers. ## Genetic Maps vs Physical Maps: The Core Comparison A genetic map is built by counting recombination events in a pedigree or mapping population. The unit is the centimorgan (cM), named after Thomas Hunt Morgan. One centimorgan corresponds to a 1 percent probability that a marker is separated from another marker by a crossover in a single meiosis. Genetic maps are inherently low resolution because recombination is rare and unevenly distributed. Two markers can sit a few kilobases apart physically and still show 1 cM of genetic distance if that spot is a recombination hotspot, or sit megabases apart and show almost no recombination if they fall in a cold region near a centromere. A physical map is built by measuring DNA directly. The unit is the base pair, kilobase (kb), or megabase (Mb), or it is an ordinal unit such as "clone 12 overlaps clone 13, which overlaps clone 14." Physical maps have no dependence on recombination frequency and therefore no dependence on sex, hotspot location, or population size. The relationship between the two is not fixed. In humans, 1 cM averages roughly 1 Mb across the genome, but this ratio varies widely by region and by sex. Recombination is suppressed near centromeres and elevated in hotspots, and female meiosis generally shows more recombination than male meiosis across much of the genome. That is why a 1 cM interval can correspond to a few hundred kilobases in one region and several megabases in another. In narrow-leafed lupin, physical-to-genetic distance ratios ranged from 11 to 109 kb per cM across the gene-rich regions studied, a spread of nearly tenfold within a single species [1]. In channel catfish, the sex-averaged genetic map spanned 3,505.4 cM, while the female-specific map spanned 4,495.1 cM and the male-specific map 2,593.7 cM, a female-to-male recombination ratio of about 1.7 to 1 [2]. Those numbers are the practical reason you cannot convert cM to Mb with a single multiplier. | Feature | Genetic map | Physical map | |--|--|--| | Unit | Centimorgan (cM) | Base pairs, kb, Mb, or clone order | | What it measures | Recombination frequency | Physical distance or overlap | | Resolution | Low, typically megabase scale | High, down to single base pairs with sequencing | | Main method | Pedigree or population genotyping | Restriction fingerprinting, STS anchoring, clone overlaps, optical mapping | | Depends on recombination | Yes | No | | Varies by sex | Yes | No | | Best use case | Trait mapping, linkage analysis, QTL detection | [Gene cloning](/blog/guides/gene-cloning), assembly validation, gap sizing, positional cloning | ## Why Physical Maps Matter Physical mapping is the bridge between a marker and a gene. A genetic map can tell you that a trait locus sits between two flanking markers. It cannot tell you how much DNA lies between them, and it cannot give you the clones you need to sequence that interval. A physical map does both. The clearest example is positional cloning. In barley, the fertility restorer locus Rfm1 was first delimited to a 10.8 cM region on chromosome 6HS. Fine mapping with 3,638 F2 plants narrowed it to a 0.14 cM interval, and a BAC physical map built from 11 isolated clones then narrowed the physical interval to 208 kb [3]. The genetic map gave the search window. The physical map gave the actual sequence to examine. Physical maps also serve as scaffolds for genome assembly. In bread wheat, whose 17 Gbp genome is polyploid and loaded with repeats, the International Wheat Genome Sequencing Consortium adopted a BAC-by-BAC strategy built on chromosome-specific physical maps [4]. Physical maps reduce the complexity problem by breaking the genome into ordered, overlapping pieces before sequencing begins. ## The Main Physical Mapping Workflows ### Restriction-Site Mapping The oldest physical mapping approach uses restriction enzymes, which cut DNA at specific recognition sequences. If you digest a chromosome with a rare-cutting enzyme, you get a set of fragments whose sizes and order define a restriction map. The distance between two cut sites is a physical distance in base pairs. Classical restriction mapping worked on single cloned fragments. Modern versions scale to whole genomes. In optical mapping, individual long DNA molecules are stretched in nanochannel arrays, digested with a restriction enzyme, and imaged. The resulting pattern of cut sites is a barcode unique to each molecule. The cattle optical map BtOM1.0 was assembled from 2,973,315 single-molecule restriction maps generated with the enzyme BamHI, spanning 2,575.30 Mb across 78 optical contigs [5]. Because the molecules are long, optical maps resolve repeat arrays that short sequencing reads collapse. In wheat chromosome 7DS, optical mapping revealed an approximately 800 kb tandem repeat array that sequencing could not resolve [4]. A related approach, RadMap, generates an ultradense whole-genome restriction map by creating hundreds of subhaploid fosmid or BAC clone pools, then sequencing the restriction site-associated DNA from each pool. Applied to Arabidopsis and human, it improved whole-genome shotgun assembly contiguity up to 15-fold, raising N50 from about 816 kb to 3.7 Mb with 98.1 to 98.5 percent scaffolding accuracy [6]. ### STS and Sequence-Tagged Site Anchoring A sequence-tagged site (STS) is a short, unique DNA sequence that can be amplified by [polymerase chain reaction](/knowledge/molecular-biology/polymerase-chain-reaction) (PCR) and that occurs exactly once in the genome. Because an STS has a known sequence, it acts as a landmark that can be detected in any DNA sample, including individual clones. STS mapping works by screening a clone library with STS primer pairs. A clone that yields a PCR product contains that STS. If two clones share an STS, they overlap at least at that site. By testing many STSs against many clones, you build a presence-absence matrix that reveals which clones overlap and in what order. The power of STSs is that they are portable. A marker developed in one laboratory can be used in another without exchanging clones. In chickpea, 245 BAC-end sequence-derived SSR markers linked the physical map to two genetic maps, allowing specific BACs to be placed near quantitative trait loci for drought tolerance and disease resistance [7]. In wheat chromosome 5BS, SSR, ISBP, and zipper markers anchored BAC clones, and 722 novel markers were developed from partial sequencing data [8]. ### Clone Fingerprinting Clone fingerprinting is the workhorse of large physical mapping projects. The idea is simple. Digest each clone with a restriction enzyme, separate the fragments by size, and record the fragment pattern. Two clones that overlap share many fragments and therefore have similar patterns. Two clones from different genomic regions share few fragments. High information content fingerprinting (HICF) uses multiple restriction enzymes and precise fragment sizing to maximize the information per clone. In chickpea, HICF of 67,483 BAC clones produced 1,174 contigs comprising 46,112 clones plus 3,256 singletons, covering a 574 Mb genome with an average contig length of 0.49 Mb [7]. In perennial ryegrass, roughly 212,000 high-information-content fingerprints were assembled into between 3,600 and 7,500 contigs depending on the software and probability thresholds used [9]. That range is a useful reminder that contig counts are threshold-dependent, not absolute. The two dominant assembly engines are Fingerprint Contig (FPC) and Linear Topology Contig (LTC). FPC uses a tolerance-based overlap rule. LTC uses a linear topology model that is more tolerant of missing bands and small sizing errors. Wheat chromosome 5BS used LTC to assemble 43,776 BAC clones into 111 scaffolds with an N50 of 3.078 Mb, covering about 99 percent of the 290 Mb arm [8]. Wheat chromosome 5A used both FPC and LTC in complementary fashion [10]. ### Contig Assembly and Minimum Tiling Paths A contig is a set of overlapping clones that together represent a contiguous stretch of DNA. Once contigs are built, the next job is to order them along the chromosome and to choose the smallest set of clones that covers each contig without gaps. That minimal set is the minimum tiling path (MTP). The MTP is what gets sequenced. Instead of sequencing every clone in a library, you sequence only the MTP clones, which overlap just enough to cover the region. In chickpea, 163 MTP clones were mapped onto eight pseudomolecules, forming 491 hypothetical contigs representing about 54 Mb of the draft genome [7]. In wheat chromosome arm 7DS, the physical map consisted of 895 contigs covering 94 percent of the estimated arm length, and anchoring assigned 73 percent of the assembly to distinct genomic positions, interconnecting 1,713 markers with ordered and sequenced MTP clones [11]. Anchoring is the step that turns an unordered contig set into a chromosome map. Anchoring methods include PCR screening of markers against clone pools, hybridization to microarrays, comparison of BAC fingerprints against in silico digests of a reference sequence, and radiation hybrid mapping. Wheat chromosome 6B integrated 689 contigs into a radiation hybrid map, determining the order and direction of 480 contigs corresponding to 87 percent of the chromosome length [12]. A high-throughput alternative sequences three-dimensional pools of MTP clones and maps the reads back to markers, which anchored 758 sequences to wheat chromosome arm 3DS in a single in silico experiment [13]. ## Worked Example: Building a Contig Step by Step Suppose you have a BAC library of 50,000 clones and a set of 200 STS markers for one chromosome. 1. **Fingerprint every clone.** Digest each BAC with two restriction enzymes, label the fragments, separate them on a capillary sequencer, and record fragment sizes. You now have 50,000 fingerprint patterns. 2. **Score overlaps.** Compare every fingerprint against every other using a tolerance rule. Two clones sharing, say, 70 percent of their fragments within a size tolerance are candidate overlaps. FPC or LTC software handles this at scale. 3. **Build contigs.** Group clones into connected components. Clones A and B overlap, B and C overlap, so A, B, and C form one contig even if A and C do not overlap directly. 4. **Screen STS markers against clone pools.** Use a three-dimensional pooling scheme so that one [PCR reaction](/knowledge/molecular-biology/pcr-reaction) tests many clones at once. A positive pool narrows the candidate clones, and a second round confirms which individual clone carries the STS. 5. **Anchor contigs.** Each STS that hits clones in two different contigs links those contigs. A genetic map position for the STS gives the contig a chromosomal coordinate. 6. **Order and orient.** Use multiple shared markers between neighboring contigs to determine order and direction. Radiation hybrid maps or optical maps can resolve ambiguous orientations. 7. **Select the minimum tiling path.** Walk along each ordered contig and pick the smallest set of clones whose overlaps cover the full length. This is what goes to sequencing. 8. **Validate.** Cross-check BAC-end sequences against neighboring clones and against the assembly. Discrepancies usually indicate a chimeric clone or a repeat-driven false overlap. The whole process is iterative. New markers close gaps. New fingerprints split or merge contigs. A physical map is a living resource, not a one-time product. ```mermaid flowchart TD [Library] --> [Fingerprint clones] [Fingerprint clones] --> [Score overlaps] [Score overlaps] --> [Build contigs] [Build contigs] --> [Screen STS markers] [Screen STS markers] --> [Anchor to chromosome] [Anchor to chromosome] --> [Order and orient] [Order and orient] --> [Select minimum tiling path] [Select minimum tiling path] --> [Sequence and validate] [Sequence and validate] --> {Gaps remain} {Gaps remain} --> [Develop new markers] [Develop new markers] --> [Screen STS markers] ``` ## Integrating Physical Maps with Genetic and Sequence Maps A physical map becomes far more useful when it is integrated with a genetic map and a sequence assembly. Integration means placing the same marker on all three maps so that positions can be translated. In channel catfish, integration of a 54,342-SNP genetic map with a BAC physical map anchored over 87 percent of physical map contigs to linkage groups, covering a physical length of 867 Mb, about 90 percent of the genome [2]. In chickpea, integration of the FPC assembly with the draft genome sequence placed 965 BACs, including 163 MTP clones, onto eight pseudomolecules [7]. In hazelnut, a high-resolution genetic map and a BAC physical map together placed the eastern filbert blight resistance locus in a single contig of three BACs, with 233 predicted genes in seven contigs within 1 cM of the locus [14]. These integrations are what make map-based cloning practical. The genetic map tells you the interval. The physical map tells you which clones span it. The sequence tells you which genes are there. ## Common Mistakes and Limitations **Chimeric clones.** A chimeric clone contains DNA from two different genomic regions joined together. This happens during library construction when two fragments ligate into one vector. A chimera creates a false overlap between two contigs that are actually far apart, and it can merge unrelated regions into one contig. The standard defense is to check overlaps with a second, independent method such as BAC-end sequencing or STS co-amplification. In the addax MHC physical map, overlaps between neighboring BACs were cross-verified by both BAC-end sequencing and co-amplification of identical PCR fragments within the overlapped region [15]. **Repetitive DNA collapsing assemblies.** Repetitive elements appear in many clones at once. Fingerprint software sees shared fragments and infers overlap, but the overlap is repeat-driven rather than positional. This collapses distinct genomic regions into a single contig and under-represents the true repeat content. Wheat is the classic case, where the repeat content hampers sequence assembly even after BAC-by-BAC sequencing [4]. Optical mapping helps because long molecules span repeats and produce unique barcodes, which is how the 800 kb tandem repeat array on wheat 7DS was resolved [4]. **Gaps at centromeres.** Centromeres are large, highly repetitive, and recombination-suppressed. Genetic maps have almost no resolution there, and clone libraries often under-represent these regions because the DNA is difficult to clone stably. Wheat chromosome 6B contigs containing parts of the nucleolus organizer region or centromere had to be characterized separately based on their radiation hybrid map positions and assembled clone sequences [12]. The addax MHC physical map contained an approximately 18 Mb gap formed by an ancient rearrangement [15]. Gaps at centromeres are expected, not exceptional. **Threshold sensitivity.** Contig counts depend on the overlap probability threshold you choose. Perennial ryegrass produced between about 3,600 and 7,500 contigs from the same fingerprint dataset depending on software and threshold [9]. Report the threshold, or the contig count is not reproducible. **Uneven coverage.** Physical maps cover gene-rich regions better than repeat-rich regions. In wheat chromosome 5A, anchored physical coverage was 75 percent for the short arm but only 53 percent for the long arm, with 64 and 48 percent of contigs ordered respectively [10]. Coverage percentages are arm-specific and should never be quoted as genome-wide. **Genetic-to-physical conversion errors.** Applying a single cM-to-Mb conversion factor across a genome is wrong. The ratio varies by region and sex, as the catfish sex-specific maps show [2] and as the lupin ratio range of 11 to 109 kb per cM demonstrates [1]. ## Quick Review - A genetic map measures recombination frequency in centimorgans and is low resolution. A physical map measures base-pair distance or clone order and is high resolution. - In humans, 1 cM averages about 1 Mb, but the ratio varies by region and sex and should never be applied as a constant. - The four main physical mapping workflows are restriction-site mapping, STS anchoring, clone fingerprinting, and contig assembly with minimum tiling paths. - FPC and LTC are the standard contig assembly engines. Contig counts depend on the overlap threshold you set. - The minimum tiling path is the smallest set of overlapping clones that covers a contig, and it is what gets sequenced. - Chimeric clones, repetitive DNA, and centromeric gaps are the three failure modes that most often corrupt a physical map. - Integration with genetic maps and sequence assemblies is what turns a physical map into a tool for gene cloning and assembly validation. ## Frequently Asked Questions ### What is the difference between a genetic map and a physical map? A genetic map orders markers by recombination frequency in centimorgans. A physical map orders markers by actual DNA distance in base pairs or by the overlap order of cloned fragments. Genetic maps are low resolution and depend on recombination. Physical maps are high resolution and do not. ### Why is 1 cM not always equal to 1 Mb? Because recombination frequency varies across the genome and between sexes. One centimorgan averages roughly 1 Mb in humans, but the ratio ranges from a few hundred kilobases in recombination hotspots to several megabases in suppressed regions near centromeres. ### What is a minimum tiling path? A minimum tiling path is the smallest set of overlapping clones that covers a contig from end to end. Sequencing only the MTP clones saves substantial cost compared with sequencing every clone in a library. ### What causes gaps in a physical map? Repetitive DNA, centromeric regions, and clone library biases are the main causes. Repeat-rich regions are hard to fingerprint uniquely, and centromeres are often under-represented in clone libraries because the DNA does not clone stably. ### How do you detect a chimeric clone? By cross-verifying overlaps with an independent method. BAC-end sequencing and co-amplification of identical PCR fragments within an overlap region are standard checks. A chimera will pass one test but fail the other. ### Can a physical map be built without cloning? Yes. Optical mapping builds a physical map from individual long DNA molecules stretched in nanochannels and digested with a restriction enzyme, with no cloning step at all. This avoids cloning bias and resolves repeat arrays that clone-based methods collapse. ## Related Articles - [How to Combine Cornell Notes with Mind Mapping for Biology](/blog/research-skills/how-to-combine-cornell-notes-with-mind-mapping-for-biology-a-hybrid-system-for-lectures-and-textbook) - [How to Create Biology Mind Maps That Actually Simplify Complex Topics](/blog/research-skills/how-to-create-biology-mind-maps-that-actually-simplify-complex-topics-a-step-by-step-guide-with-exam) - [The Best Tools for Creating Biology Concept Maps](/blog/research-skills/the-best-tools-for-creating-biology-concept-maps-a-comparison-of-free-and-paid-mind-mapping-software) - [Choosing the Right Read Mapper for Bacterial Genomics](/knowledge/bacteria/general/choosing-the-right-read-mapper-for-bacterial-genomics-a-comparative-review-of-bwa-bowtie2-and-minima) - [Is Environmental Science a Physical Science?](/blog/careers/is-environmental-science-a-physical-science) - [Physical Therapy Degree](/blog/careers/physical-therapy-degree) - [UCSC Genome Browser: Beginner Tutorial and Tips](/knowledge/bioinformatics/ucsc-genome-browser-beginner-tutorial-and-tips) - [Central Sulcus: Location, Function and Motor Map](/knowledge/veterinary-medicine/veterinary-anatomy-physiology/central-sulcus-location-function-and-motor-map) - [MAP and TAU in Bayesian Statistics](/blog/research-skills/map-and-tau-in-bayesian-statistics) ## Sources 1. [Remnants of the Legume Ancestral Genome Preserved in Gene-Rich Regions: Insights from Lupinus angustifolius Physical, Genetic, and Comparative Mapping.](https://pubmed.ncbi.nlm.nih.gov/25620837/) 2. [Construction of a high-density, high-resolution genetic map and its integration with BAC-based physical map in channel catfish.](https://pubmed.ncbi.nlm.nih.gov/25428894/) 3. [High-resolution genetic mapping and physical map construction for the fertility restorer Rfm1 locus in barley.](https://pubmed.ncbi.nlm.nih.gov/25412992/) 4. [BioNano genome mapping of individual chromosomes supports physical mapping and sequence assembly in complex plant genomes.](https://pubmed.ncbi.nlm.nih.gov/26801360/) 5. [A clone-free, single molecule map of the domestic cow (Bos taurus) genome.](https://pubmed.ncbi.nlm.nih.gov/26314885/) 6. [Whole-Genome Restriction Mapping by "Subhaploid"-Based RAD Sequencing: An Efficient and Flexible Approach for Physical Mapping and Genome Scaffolding.](https://pubmed.ncbi.nlm.nih.gov/28468906/) 7. [Integrated physical, genetic and genome map of chickpea (Cicer arietinum L.).](https://pubmed.ncbi.nlm.nih.gov/24610029/) 8. [Features of the organization of bread wheat chromosome 5BS based on physical mapping.](https://pubmed.ncbi.nlm.nih.gov/29504906/) 9. [Integrating a newly developed BAC-based physical mapping resource for Lolium perenne with a genome-wide association study across a L. perenne European ecotype collection identifies genomic contexts associated with agriculturally important traits.](https://pubmed.ncbi.nlm.nih.gov/30715119/) 10. [Physical Mapping of Bread Wheat Chromosome 5A: An Integrated Approach.](https://pubmed.ncbi.nlm.nih.gov/33228274/) 11. [Integrated physical map of bread wheat chromosome arm 7DS to facilitate gene cloning and comparative studies.](https://pubmed.ncbi.nlm.nih.gov/29526810/) 12. [A high-resolution physical map integrating an anchored chromosome with the BAC physical maps of wheat chromosome 6B.](https://pubmed.ncbi.nlm.nih.gov/26265254/) 13. [High-throughput physical map anchoring via BAC-pool sequencing.](https://pubmed.ncbi.nlm.nih.gov/25887276/) 14. [High-Resolution Genetic and Physical Mapping of the Eastern Filbert Blight Resistance Region in 'Jefferson' Hazelnut (Corylus avellana L.).](https://pubmed.ncbi.nlm.nih.gov/28724074/) 15. [A high-density BAC physical map covering the entire MHC region of addax antelope genome.](https://pubmed.ncbi.nlm.nih.gov/31185912/)