A physical map of the mouse genome
Simon G. Gregory*, Mandeep Sekhon†, Jacqueline Schein‡, Shaying Zhao§, Kazutoyo Osoegawak, Carol E. Scott*, Richard S. Evans*,
Paul W. Burridge*, Tony V. Cox*, Christopher A. Fox*, Richard D. Hutton*, Ian R. Mullenger*, Kimbly J. Phillips*, James Smith*, Jim Stalker*,
Glen J. Threadgold*, Ewan Birney{, Kristine Wylie†, Asif Chinwalla†, John Wallis†, LaDeana Hillier†, Jason Carter†, Tony Gaige†,
Sara Jaeger†, Colin Kremitzki†, Dan Layman†, Jason Maas†, Rebecca McGrane†, Kelly Mead†, Rebecca Walker†, Steven Jones‡,
Michael Smith‡, Jennifer Asano‡, Ian Bosdet‡, Susanna Chan‡, Suganthi Chittaranjan‡, Readman Chiu‡, Chris Fjell‡, Dan Fuhrmann#,
Noreen Girn‡, Catharine Gray‡, Ran Guin‡, Letticia Hsiao‡, Martin Krzywinski‡, Reta Kutsche‡, Soo Sen Lee‡, Carrie Mathewson‡,
Candice McLeavy‡, Steve Messervier‡, Steven Ness‡, Pawan Pandoh‡, Anna-Liisa Prabhu‡, Parvaneh Saeedi‡, Duane Smailus‡,
Lorraine Spence‡, Jeff Stott‡, Sheryl Taylor‡, Wesley Terpstra‡, Miranda Tsai‡, Jill Vardy‡, Natasja Wye‡, George Yang‡, Sofiya Shatsman§,
Bola Ayodeji§, Keita Geer§, Getahun Tsegaye§, Alla Shvartsbeyn§, Elizabeth Gebregeorgis§, Margaret Krol§, Daniel Russell§, Larry Overton§,
Joel A. Malek§, Mike Holmes§, Michael Heaney§, Jyoti Shetty§, Tamara Feldblyum§, William C. Nierman§, Joseph J. Catanesek,
Tim Hubbard*, Robert H. Waterston†, Jane Rogers*, Pieter J. de Jongk, Claire M. Fraser§, Marco Marra‡, John D. McPherson†
& David R. Bentley*
* The Wellcome Trust Sanger Institute, Hinxton, Cambridge CB10 1SA, UK
† Genome Sequencing Center, Washington University School of Medicine, St Louis, Missouri 63108, USA
‡ Genome Sciences Centre, British Columbia Cancer Agency, Vancouver, British Columbia V5Z 4E6, Canada
§ The Institute for Genomic Research, Rockville, Maryland 20850, USA
k Children’s Hospital Oakland Research Institute, Oakland, California 94609, USA
{ EMBL—European Bioinformatics Institute, Hinxton, Cambridge CB10 1SD, UK
# Department of Electrical Engineering, Washington University, St Louis, Missouri 63130, USA
...........................................................................................................................................................................................................................
A physical map of a genome is an essential guide for navigation, allowing the location of any gene or other landmark in the
chromosomal DNA. We have constructed a physical map of the mouse genome that contains 296 contigs of overlapping bacterial
clones and 16,992 unique markers. The mouse contigs were aligned to the human genome sequence on the basis of 51,486
homology matches, thus enabling use of the conserved synteny (correspondence between chromosome blocks) of the two
genomes to accelerate construction of the mouse map. The map provides a framework for assembly of whole-genome shotgun
sequence data, and a tile path of clones for generation of the reference sequence. Definition of the human–mouse alignment at this
level of resolution enables identification of a mouse clone that corresponds to almost any position in the human genome. The
human sequence may be used to facilitate construction of other mammalian genome maps using the same strategy.
The recent revolution in large-scale analysis of genomes has already
yielded near-complete DNA sequences of a diverse range of organisms
including bacteria, a yeast, a worm, a fly and humans1–7
. The
sequence of the mouse genome will have a huge impact on
biological research and human health. It will provide critical
information and reagents for use in mouse experimental models.
It will become possible to unravel the mechanisms of complex
mammalian biological processes and human disease. The mouse
genome sequence will provide the first opportunity to compare the
complete sequence and organization of the human genome with
that of another mammal. It will aid discovery and annotation of
gene structures and other functionally important sequences in both
genomes.
Assembly of each large genome sequence to date has been
underpinned by production of a comprehensive map of overlapping
large-insert bacterial clones (for example, cosmids8
or bacterial
artificial chromosome (BAC) clones9
)10–13
for sequencing, and, in
some cases, for integration with whole-genome shotgun sequence
data (refs 5, 7; see also review in ref. 14). The clones provide
substrates for finishing each section of the genome sequence
separately, making it easier to eliminate errors and to achieve a
final accuracy of more than 99.99% (refs 14–16). The study of other
large genomes requires physical maps of a similar standard for
sequencing and biological studies.
Comparisons between genomes reveal homologous sequences,
for example within individual genes, that reflect their common
evolutionary origin and subsequent conservation. Similarity
between genomes is also evident at the level of long-range sequence
organization. A ‘conserved segment’ (also called a ‘conserved
linkage’) refers to a region where the order of multiple genes on a
single chromosome segment (and hence the linkage between them)
is the same in both species. A ‘conserved synteny’ refers to a region
where the chromosomal location of multiple genes is conserved, but
not necessarily their precise order17–19
. In general, the degree of
similarity at all levels is higher between species that diverged more
recently from a common ancestor. The long-range organization of
the mouse and human genomes is very similar, and may comprise
about 180 conserved syntenic regions6,17
. Detailed studies in some
smaller regions suggest that precise gene order has been conserved,
confirming the existence of conserved segments20,21
. The similarity
in sequence organization between these two genomes suggests that
the human genome can be used as a framework to map the mouse
genome20,21
, and possibly other mammalian genomes as well.
Alignment of the different genome maps over their entire length
would simultaneously define the syntenic relationship between
them at a new level of resolution, and accelerate the process of
constructing the mouse clone map.
Using the human sequence as a framework
We constructed a physical clone map of the mouse genome in two
phases. In the first phase, we compared restriction digest patterns
(‘fingerprints’) of 305,716 BAC clones22
. We identified overlaps
between clones on the basis of similarity between fingerprints and
used this information to construct 7,587 contigs of overlapping
clones. In parallel, we determined sequences at the ends of the
mouse DNA fragments cloned in the BACs (that is, BAC end
sequences, or BESs). We aligned the mouse BAC contigs to the
human genome sequence by identifying matches between human
sequence and mouse BESs. We extended and joined contigs where
possible after re-examining the fingerprint data. Overlapping fin-
articles
NATURE | VOL 418 | 15 AUGUST 2002 | www.nature.com/nature 743© 2002 NaturePublishing Group
gerprints were required to support each join, while juxtaposition of
the mouse clone contigs along the human genome greatly accelerated
the manual editing process. This phase resulted in generation
of a human–mouse homology clone map. In the second phase, we
used a set of independently mapped mouse markers (available in
existing genetic and radiation hybrid maps of the mouse) to
position the BAC contigs in the mouse genome. After further
manual contig editing was carried out, we generated a mouse
clone map comprising 296 contigs.
An illustration of the human–mouse homology clone map
produced in the first phase is shown in Fig. 1. Within a 1.6megabase
(Mb) region of human chromosome 6 (6q16.1), 11 of
the 15 segments of human sequence match to 29 of the BESs within
a mouse BAC contig. This contig is located on mouse chromosome
4 (Mmu4). The syntenic relationship between the two genomes in
this area was previously indicated by the proximity of the two loci
CNR1 and GABRR1 (human) and their corresponding homologues
Cnr1 and Gabrr1 (mouse) (also shown in Fig. 1). In our analysis
here, a further 29 homologous crosslinks between these two loci are
added.
We found 51,486 homologous crosslinks between the two genomes
during this analysis, which corresponds to one match per
54 kb on average (assuming a 2.8-Gb mouse genome; see Methods).
These matches were derived from a complete set of 453,962 BESs.
Some 73% of the BESs contained more than 100 base pairs (bp) of
contiguous unique sequence; we excluded the rest from the analysis.
In all, 1,326 of the BESs matched human chromosome 20 sequence.
We found that 19% of them matched putative coding regions (163
matches to known genes, 80 to new genes, 8 to putative genes and 1
to a pseudogene). The rest were matches to other conserved
sequences in introns (32%) and intergenic regions (49%).
Human–mouse sequence homologies in non-coding regions have
also been observed in other studies20,23,24
. From this analysis it is
clear that non-coding homologies contributed significantly to the
alignment.
Of the clones in the human genome tile path, 88% are collinear
with the mouse BAC map (Table 1). For individual human
chromosomes, coverage by aligned mouse contigs exceeds 80% on
all except chromosome 19 (61%) and the Y chromosome (0%).
Lower coverage of these human chromosomes can be accounted for
by several factors. First, we used only BES matches with unique
locations in the human tile path for the alignment. BESs matching
regions of human sequence containing dispersed low-copy repeats
were filtered out of the analysis. Such regions are known on human
chromosome 19 (notably including zinc-finger-containing genes
among others21,25
). Second, a significant fraction of chromosome 19
clones (49 in all) are included in the total in Table 1 but contain
pericentromeric repetitive sequence, and any BES matches to them
were excluded. A further 118 clones on chromosome 19 were
covered by contigs aligned in our study, but we did not score
them in Table 1 because they contained no high-scoring matches to
BESs. This is reflected in the low overall density of BES matches on
Mmu4Mbc
A
B
C
D
E
97.7
97.9
98.5
98.1
98.2
98.4
98.3
99.0
99.1
99.3
99.2
98.6
98.7
98.9
98.8
97.8
98.0
ContigsaHsa6 Contigs
6.8
7.4
7.9
6.9
7.2
7.3
7.8
7.5
7.6
7.7
7.0
7.1
8.0
6.7
D4Mit137
D4Mit94
Cnr1CNR1
Gabrr1GABRR1
Mb
1
1
2
3
4
5
1
2
3
1
2
3
4
5
6
7
1
2
3
2
1
b
25
24
23
22
21.3
21.2
21.1
12
12
13
14
15
16
21
22
23
24
25
26
27
11
p
q
Figure 1 Construction of human–mouse homology clone map. Alignment between part of
human chromosome 6 (Hsa6) and mouse chromosome 4 (Mmu4). Clone contigs are
shown as blue and pale pink boxes, respectively. A 1.6-Mb interval is enlarged, showing
part of Hsa6q16.1 aligned to a 1.3-Mb mouse BAC contig. The accompanying megabase
scale for Hsa6 starts from one end of the chromosome (ptel, 0 Mb; qtel,172 Mb), whereas
the megabase scale shown for the mouse contig starts at one end of that contig. a, Human
accessions. Filled and clear boxes denote accessions with and without matches to BESs,
respectively. b, Matches between human sequence and BESs. Black lines denote highhomology
BLAST matches and dashed lines indicate medium-homology BLAST matches
(see Methods for details). c, Mouse BAC clone contig. All clones having at least one BES
that matches a human accession are shown in red (black and white squares at the ends of
clones denote BESs with and without matches to human accessions, respectively). Black
lines represent a subset of the other clones in the contig (the rest have been omitted for
clarity, but can be viewed in Supplementary Information). Previous alignment of the two
human loci CNR1 and GABRR1 with their mouse homologues can be viewed in the NCBI
versus MGD alignment (http://www.ncbi.nlm.nih.gov/Homology). Another human locus
(RNGTT ) lies between CNR1 and GABRR1 in the human sequence, and its murine
homologue (Rngtt ) is present in the NCBI versus MGD alignment; however, Rngtt has not
been placed in the mouse clone map yet. D4Mit94 and D4Mit137 are additional mouse
markers that have been placed in the clone map by this study.
articles
NATURE | VOL 418 | 15 AUGUST 2002 | www.nature.com/nature744 © 2002 NaturePublishing Group
this chromosome (Table 1) and may be the result of chromosomespecific
low-copy repeats, as observed previously21,25
. We were not
able to align any mouse contigs to the human Y chromosome. Two
reasons may account for this. First, much of the human Y sequence
is repetitive (with Y-specific, X–Y and autosome–Y homologies),
and mouse BES matches to these regions would have been excluded.
Second, only 39% of the BESs were derived from a male mouse
library, and we estimate that only about 0.4% of the BESs were
therefore derived from the mouse Y chromosome.
Of the total coverage of the mouse BAC map (in 211 contigs),
97% (2,658 Mb) is aligned to the human genome sequence. The
other 3% (84 Mb, in 85 contigs) may not have been aligned at this
stage for several reasons. First, the BESs may not find matches in the
human sequence because the human sequence is incomplete. In a
recent analysis, 3% of a curated set of 10,212 full-length human
complementary DNAs (taken from the set in the ‘reference sequence
project’, http://www.ncbi.nlm.nih.gov) were found to have no
match in the human sequence (using BLAT26
to identify matches
with more than 95% identity between cDNA and the human
genome sequence (NCBI Build 28); J.Kent and D. Haussler, personal
communication). This fraction is sufficient to account for all of the
unplaced mouse contigs. Second, most of the unplaced contigs are
Table 1 Summary statistics of human–mouse homology clone map
Human chromosome
number
Human sequence
available (Mb)
Human–mouse sequence
matches
Number of clones in
human tile path
Clones in human tile path that
are covered in mouse
Total Per Mb Number Per cent
...................................................................................................................................................................................................................................................................................................................................................................
1 239 4,888 20.5 2,179 2,060 95
2 242 4,783 19.8 1,951 1,675 86
3 204 3,667 18.0 1,700 1,493 88
4 189 2,747 14.5 1,612 1,510 94
5 182 2,892 15.9 1,973 1,682 85
6 176 3,264 18.5 1,800 1,666 93
7 161 2,957 18.4 1,518 1,342 88
8 143 2,310 16.2 1,210 1,040 86
9 114 2,505 22.0 963 883 92
10 140 2,531 18.1 1,132 1,014 90
11 139 2,606 18.7 1,134 1,013 89
12 137 2,206 16.1 1,022 943 92
13 99 1,318 13.3 851 798 94
14 90 2,103 23.4 655 642 98
15 82 1,721 21.0 671 573 85
16 81 1,470 18.1 726 666 92
17 81 1,585 19.6 656 612 93
18 79 1,226 15.5 616 564 92
19 46 507 11.0 841 512 61
20 61 1,326 21.8 632 608 96
21 34 507 14.9 475 432 91
22 34 439 12.7 345* 279 81
X 149 1,925 12.9 1,555 1,276 82
Y 26 0 0.0 208 0 0
All 2,928 51,483 17.5 26,425 23,283 88
...................................................................................................................................................................................................................................................................................................................................................................
*For chromosome 22, the finished sequence was divided into abutting 100-kb segments for this analysis.
Table 2 Summary statistics of mouse physical clone map by chromosome
Size estimate (Mb) Total contigsMouse chromosome
number
Old* New Number Coverage (Mb)
Percentage of old
size estimate
Percentage of new
size estimate
...................................................................................................................................................................................................................................................................................................................................................................
1 216 196 10 211 98 108
2 209 190 13 186 89 98
3 180 163 8 173 96 106
4 177 160 10 145 82 90
5 170 154 24 156 92 101
6 166 151 16 157 95 104
7 156 141 14 143 92 101
8 149 135 12 140 94 104
9 144 131 15 127 88 97
10 145 131 9 135 93 103
11 142 129 6 107 75 83
12 146 132 8 126 86 95
13 131 119 11 132 101 111
14 134 121 8 130 97 107
15 122 111 7 108 89 98
16 114 103 7 101 88 97
17 116 105 6 95 82 91
18 116 105 6 96 83 91
19 82 74 4 65 80 88
X 187 170 30 132 71 78
Y 86 78 4 14 17 18
Subtotal 228 2,679 87 96
Unlocalized 68 60
Total 3,088 2,800 296 2,739 89 98
...................................................................................................................................................................................................................................................................................................................................................................
*Taken from ref. 47.
articles
NATURE | VOL 418 | 15 AUGUST 2002 | www.nature.com/nature 745© 2002 NaturePublishing Group
short and contain few BESs. The production of additional mouse
sequence in these contigs may allow identification of homology
crosslinks, and hence alignment of the contigs to the corresponding
region in the human genome. Third, there may be insufficient
homology between the two genomes in some regions.
The mouse clone map
The mouse genome comprises 19 autosomes plus the two sex
chromosomes, X and Y. In contrast to the human genome, all the
mouse chromosomes are acrocentric (that is, the centromere is at or
near one end). To construct the clone map for each mouse chromosome,
we made use of the available genetic map (http://wwwgenome.wi.mit.edu/cgi-bin/mouse/index)
of 6,559 gene-specific or
simple sequence length polymorphism (SSLP) markers27–29
, and also
a radiation hybrid (RH) map30
of 13,558 markers, which included a
subset of 2,446 of the markers previously ordered in the genetic
map. This provided a total unique set of 18,379 markers. We placed
48% (8,789) of these markers within the mouse BAC contigs using
the BESs or matches to available mouse genomic sequence, or by
hybridization of markers to arrayed BAC clones31–33
. We then added
another 8,203 markers to the map by hybridization to the BACs. As a
result, we were able to assign a unique position and orientation on a
mouse chromosome to 203 contigs (containing 2,631 Mb, or 96% of
the map coverage), and a chromosome assignment (but not a
position) to 25 more contigs (containing a total of 47 Mb). A
further 68 contigs (containing 60 Mb) have no chromosomal
assignment. A summary of the mouse map by chromosome is
shown in Table 2.
Most mouse BAC contigs contained multiple mouse markers
(average 57 markers per contig). We compared the order of markers
within the contigs to the order suggested previously by the genetic
and RH data. Overall, there was excellent agreement between the
order of markers in each map, as illustrated for example on mouse
chromosome Mmu2 (see Fig. 2; and Supplementary Information
for all chromosomes). There was also concordance of the marker
order in the clone map with that of the European Union RH map
(http://www.genoscope.cns.fr/externe/English/Projets/Projet_
ZZZ/rhmap.html). The availability of integrated map information
permitted assessment of the remaining 27 minor conflicts. We
found two conflicts between the clone map and both the genetic
and RH maps. The other 25 were conflicts solely with the RH map,
and occur in regions where the genetic map offers limited or no
resolution. For example, we inspected the RH data in the 80–90-Mb
region of Mmu2 (Fig. 2) and found that the number of obligate
breaks between the framework markers can be reduced if these
markers are arranged according to the order in the clone map. We
made the same observation in all eight cases examined in detail
(see Supplementary Information). In general, it appears that the
conflicts are caused by local inversions owing to a few (typically one
or two) suboptimally located framework markers in the RH map.
This in turn causes the order of all other markers previously placed
in the RH map relative to that framework marker to be in conflict
with their order in the clone map.
We carried out an additional check of contig assembly by
comparing ‘virtual’ fingerprints calculated from finished mouse0.3
0.7
0.1
ND
1.3
0.4
a
b
c
d
A B C D E
1 2 3 1 2 3 1 2 3 4 5 1 3 1
F G H
2 2 3 4
0.7
1.0
1.0
ND
ND
Chromosome position (Mb)
Geneticposition(cM)
Radiationhybridposition(cR)
0
500
1,000
1,500
2,000
2,500
0 20 40 60 80 100 120 140 160 180
0
20
40
60
80
100
120
Mmu2
ND
Figure 2 Comparison of radiation hybrid and genetic maps to the physical map of mouse
chromosome 2. a, Radiation hybrid (pink) and genetic (dark blue) markers are plotted
against their respective positions in the physical map of chromosome 2. b, Mouse
chromosome 2 ideogram. c, d, Mouse BAC contigs (c) and gap sizes (d), estimated from
bridging human sequence as described in the text. cR, centiRays; cM, centimorgans; ND,
not determined.
Mmu1
Mmu4
Mmu17
Mmu10
7
2
3
5
9
8
11
13
14
15
16
17
18
19
20
4
6
12
1
Mmu9
2
Mmu13
10
Hsa6
Figure 3 Conserved segments between human chromosome 6 (Hsa6) and the mouse
genome. Blue bars in the centre of the chromosome ideogram represent the current
extent of contig coverage in the human map. Numbered accession blocks correspond to
Table 3 and denote conserved synteny with the mouse chromosomes. Dotted lines joining
coloured blocks between the two species indicate conserved syntenic blocks inverted with
respect to each other relative to the orientation of each chromosome as drawn.
articles
NATURE | VOL 418 | 15 AUGUST 2002 | www.nature.com/nature746 © 2002 NaturePublishing Group
sequence with the experimental fingerprints in the database. This
was possible for 762 clones, each with more than 100 kb of finished
sequence available. All virtual fingerprints were matched with high
confidence to an experimentally derived fingerprint in the database,
thus confirming the accuracy of the fingerprint data. Some 97% of
them also matched to the expected position in contigs, confirming
the assembled BAC map. The remaining 3% (20 clones) matched to
other unique locations. These can be attributed to errors in tracking
clone names during the project, and are generally resolved later by
checking the overlaps between clones using sequence information as
it becomes available.
Assuming the mouse genome is about 2.8 Gb in total (see
Methods), coverage of the mouse genome in mapped BACs is
virtually complete: 296 contigs of average size 9.3 Mb cover an
estimated 2,739 Mb. The estimated coverage of each mouse
chromosome by BAC contigs is listed in Table 2. Four contigs
were placed on the mouse Y chromosome. The low coverage is
partly due to the lack of previously mapped markers, and partly
because only part of the fingerprinted BAC collection was derived
from a male mouse. A number of the unplaced contigs contain
clones derived solely from the male mouse BAC library, and may
also map to the mouse Y chromosome when more data is available.
The map continues to be updated, and the current version can be
viewed at http://www.ncbi.nlm.nih.gov/genome/guide/mouse
(choose ‘MapViewer’) or at http://www.ensembl.org/Mus_muscu
lus/cytoview (use the ‘Jump to chr.’ box to select chromosome; if
searching specific features, select ‘cytoview’ option to view the
map). The archive of the map at the time of publication is available
as Supplementary Information. The level of continuity of the mouse
clone map is similar to that of the current human genome map,
which comprises 279 contigs covering approximately 2,928 Mb (as
of March 2002; International Human Genome Sequencing Consortium,
I. Vastrik, personal communication).
There are 275 gaps in the mouse clone map. Many of these gaps
occur where there is a break in synteny between the two genomes.
No direct measurement of the existing gaps has yet been attempted
(for example by fluorescent in situ hybridization to chromosomes or
DNA fibres). However, 83 gaps in the mouse map are bridged by
human contigs. The total length of human sequence across these
gaps is 51.9 Mb. If we assume collinearity between the mouse and
human genomes across these gaps, then the average size of each gap
in the mouse map is 0.63 Mb. On mouse chromosome 2, for
example, 9 of the 12 gaps between contigs are bridged by segments
of the human sequence ranging from 0.3 to 1.3 Mb (see Fig. 2). More
practically, the tracts of human sequence that bridge gaps in the
mouse map may provide a source of probes (for example, selected
from annotated exons in the human sequence) to screen for new
mouse clones and thus assist in closing the gaps. Conversely, mouse
clones that are collinear with gaps in the human tile path may be
used to assist closure of the human map. Our study indicates that
132 of the 244 gaps in the human map (March 2002 release) are
bridged by mouse contigs and might be closed in this way.
Human–mouse homology maps
Alignment of the human and mouse clone maps using 51,486
homology crosslinks allowed us to define the segments that are
conserved in both genomes at a new level of resolution. In many
cases we were able to locate the boundary between adjacent
segments to within one or a few clones. For example, the sequence
of human chromosome 6 contains 1,800 sections that overlap to
form eight contigs. We numbered each section arbitrarily from 1
(nearest the telomere of the short arm) to 1,800 (nearest the longarm
telomere) for the purpose of this analysis (Table 3, ‘ptel–qtel’
numbers). We detected 220 high or medium stringency matches
between a 9-Mb region of human 6p25.3 (segment 6.1, represented
by accessions 3–97) and a region of about 9.9 Mb in a mouse BAC
contig mapped to Mmu13A3 (Table 3 and Fig. 3). In addition, we
found that the order of all 220 matches is consistent between the two
maps, supporting the possibility that this is a conserved segment.
The next segment (segment 6.2) starts in human accession 99 and
extends as far as accession 207. This segment is defined by 303
matches of consistent order with a contig on Mmu13. Segments 6.1
and 6.2 are discontinuous in the mouse. The next segment (6.3) is
aligned to another discontinuous region of Mmu13, whereas the
fourth block (6.4) on human 6p is syntenic with a different mouse
chromosome, Mmu17 (Table 3 and Fig. 3). In total, 20 conserved
segments were identified along human chromosome 6, which
subdivide the nine regions of conserved synteny (Fig. 3; see also
Supplementary Information and updated views at http://www.ensembl.org/Homo_sapiens/syntenyview).
The results of this analysis
can also be viewed with respect to the mouse chromosomes. For
example, mouse chromosome 11 contains five regions of conserved
synteny with the human genome, which are subdivided into 21
conserved segments (Fig. 4; see also Supplementary Information
and updated views at http://www.ensembl.org/Mus_musculus/syn-
tenyview).
Overall, we defined 288 conserved segments contained within 167
regions of conserved synteny (Fig. 5). This compares with a previous
Table 3 Conserved segments between human chromosome 6 and mouse
Segment Human ptel–qtel
number*
GenBank accession
numbers
Human segment
size (Mb)
Human–mouse
homology hits
Hits per
Mb
Mouse chromosome
number
Mouse segment
size (Mb)
...................................................................................................................................................................................................................................................................................................................................................................
6.1 3–97 AL365272–AL031785 9.0 220 24 13 9.9
6.2 99–207 AL031904–AL022726 10.3 303 29 13 9.0
6.3 209–302 AL008627–AL049543 9.5 146 15 13 9.9
6.4 310–374 AL022727–Z97183 5.6 89 16 17 3.4
6.5 376–432 AL161903–AL035690 5.1 116 23 17 4.4
6.6 434–453 AL136087–AL031778 1.8 37 21 17 1.9
6.7 458–537 AL139331–AL390247 6.5 178 27 17 6.6
6.8 560–582 AL360175–AL109918 2.7 70 26 1 3.1
6.9 582–613 AL109918–AL589796 3.1 57 18 9 2.9
6.10 620–771 AL031779–AL365232 12.4 204 16 1 13.2
6.11 773–900 AL603910–AL136082 11.1 257 23 9 10.9
6.12 918–998 AL096817–AL513186 7.0 174 25 4 7.5
6.13 1,005–1,031 AL607077–AL589740 2.9 45 16 4 2.7
6.14 1,033–1,044 Z84482–AL390959 1.1 24 22 4 1.7
6.15 1,048–1,120 AL080285–AL645528 6.8 120 18 10 8.3
6.16 1,121–1,227 AL355586–AL121953 9.2 157 17 10 10.4
6.17 1,229–1,297 AL354716–AL512283 6.2 141 23 10 7.0
6.18 1,301–1,579 AL591428–AL078581 26.1 616 24 10 27.7
6.19 1,584–1,637 AL355497–AL591419 3.9 65 17 10 4.1
6.20 1,639–1,797 AL591499–AL031259 15.0 246 16 17 11.4
Total 155.3 3,265 21 156.0
...................................................................................................................................................................................................................................................................................................................................................................
*Numbered arbitrarily from 1 (nearest the telomere of the short arm) to 1,800 (nearest the long-arm telomere).
articles
NATURE | VOL 418 | 15 AUGUST 2002 | www.nature.com/nature 747© 2002 NaturePublishing Group
analysis using 2,997 markers, in which 183 conserved segments were
detected6
. The existence of multiple adjacent (but discontinuous)
segments within the same syntenic region (for example segments
6.1–6.3 and 6.15–6.19 in Table 2 and Fig. 3) supports the hypothesis
that a significant number of intrachromosomal inversions, in
addition to interchromosomal rearrangements, contributed to
formation of the present-day structures from the chromosomes of
a common ancestor. Our results are consistent with previous
observations for several chromosomal regions34–37
, including the
recent high-resolution studies on human chromosomes 19 (refs 21,
38), 21 (ref. 39) and 7 (ref. 20). In the latter study, for example,
comparative analysis using a high density of markers resulted in
detection of 20 conserved segments within 16 regions of conserved
synteny (compared with 25 and 15, respectively, in the present
study). Small conserved segments will inevitably escape detection in
the absence of sufficient homology crosslinks. For example, the
Mmu2–Hsa7 segment reported in the previous study was estimated
to be probably less than 100 kb (ref. 20), and was not detected in
our study here. Anecdotal evidence from detailed comparisons
of finished sequence between the mouse and human chromosomes
indicates the presence of very small regions (as little as 1 kb)
where conserved order is disrupted, either by recent duplications
or insertion of unrelated sequence23,34,39,40
. Ultimately, the alignment
of finished sequences of the two genomes will be required to
define the exact number and boundaries of every conserved
segment.
Future work
Physical clone maps of complex genomes provide an essential
framework to produce complete genome sequences. The clones
also provide a source of well-characterized material for experimental
studies. In the absence of large-scale sequencing programmes
(and these may not be expected for a number of genomes in the
foreseeable future), the availability of an accurate clone map will
allow targeted sequencing of regions of specific interest. The
alignment of maps of multiple genomes will also permit rapid
investigation of orthologous gene sequences for structural and
functional comparisons.
The first genomes to be mapped and sequenced in their entirety
have been the result of enormous and expensive efforts. A hallmark
of these studies has been the quality of the product, which has
benefited from cross-validation between independently constructed
maps for long-range information, and from the high accuracy of
finished sequence. In our study, the availability of the human
genome sequence has been exploited for faster characterization of
the mouse genome. This approach will be applicable to many other
mammals, and may also work for more distantly related vertebrates.
Three factors contribute to the success of the mapping strategy: (1)
good coverage of the genome in large contigs after the initial
fingerprinting; (2) high-quality sequence cross-matches between
most of these contigs and the reference sequence; and (3) sufficient
independently mapped markers to order and orient all the contigs
in the new map. Future projects would benefit from a similar depth
of coverage in BACs (at least 15-fold) to ensure good representation
of the genome in long contigs. Some savings might be made in the
number of BESs required for the mapping alignment. One way to
reduce the cost of the project would be to select clones from preexisting
contigs before sequencing the insert ends. However, this
would reduce the number of sequence anchor points in the final
map with consequent disadvantages (see below). For genomes more
closely related than human and mouse, a higher proportion of the
BESs would be expected to match the reference sequence (compared
with 16% for the human–mouse study), which would allow
reduction in the number of BESs generated without reduction of
cross-matches.
The physical clone map of the mouse has provided 98% coverage
of the genome in BAC contigs of average size more than 9 Mb. From
the more than 300,000 BACs in the map, a complete tiling path of
clones is being selected for production of the finished sequence. So
Hsa5
Hsa17
Hsa2
Hsa7
Hsa22
Mmu11
Figure 4 Conserved segments between mouse chromosome 11 and the human genome.
The pale green bars in the centre of the chromosome ideogram represent the current
extent of contig coverage in the mouse map. Dotted lines between the two species of
same coloured blocks indicate conserved syntenic blocks inverted with respect to each
other relative to the orientation of each chromosome as drawn.
Figure 5 Homology maps of the mouse and human genomes. a, The mouse–human
homology map. Bars in the centre of each mouse chromosome ideogram denote the
ordered and oriented mouse contigs determined by this study. The coloured blocks
adjacent to each mouse chromosome indicate discrete regions where marker order is
conserved between the two genomes. Colours correspond to specific human
chromosomes in b. Positions of human telomeres are located within mouse contigs by
arrows in the colour of their corresponding chromosome: up arrow indicates the p
telomere and down arrow indicates the q telomere. The X-chromosome relationship is
represented by arrows drawn in the direction Xp to Xq from the human. b, The human–
mouse homology map. Bars in the centre of human chromosome ideograms denote clone
contigs in the human genome. Adjacent coloured blocks indicate discrete regions where
marker order is conserved with the mouse. Colours correspond to specific mouse
chromosomes as in a. An interactive display of these figures is available (http://
www.sanger.ac.uk/Projects/M_musculus/publications/fpcmap-2002). Regularly
updated synteny information is also available (http://www.ensembl.org/Mus_musculus/
syntenyview and http://www.ensembl.org/Homo_sapiens/syntenyview).
Q
articles
NATURE | VOL 418 | 15 AUGUST 2002 | www.nature.com/nature748 © 2002 NaturePublishing Group
far, about 2.8% of the mouse genome sequence is available in this
form (http://www.ncbi.nlm.nih.gov), and completion is scheduled
for 2005. The mouse map contains more than 450,000 BESs
(covering 220 Mb, or 7.8%, of the mouse genome), most of which
contain sufficient unique sequence to anchor the current mouse
whole genome sequence assemblies of 40 million reads (,6 £
coverage; http://trace.ensembl.org). This combined resource will
provide a comprehensive working draft of the mouse genome in the
public domain for immediate use. For example, the January 2002
release of this product included 20,652 annotated transcripts
(http://www.ensembl.org/Mus_musculus/stats). The same
sequence is also assisting the identification of genes on the basis
of sequence conservation41
where no human cDNA is available.
Detailed comparison between human and mouse sequences will
therefore provide a comprehensive catalogue of genes (and orthologous-pair
relationships where they exist) for comparative functional
and genetic studies. It will also be possible to identify the
evolutionary events at the sequence level that have occurred since
divergence of the two species. A
Methods
Map construction
Fingerprints of RPCI-23 and RPCI-24 mouse BAC clones22
were generated by separation
of HindIII restriction-enzyme digest fragments on 1.2% agarose gels42
. Gel images were
acquired with Image software (http://www.sanger.ac.uk/Software/Image) and restriction
fragments identified using BANDLEADER software (D. Fuhrmann et al., unpublished).
Fingerprints were assembled within the program FPC43
under high stringency conditions
at a probability of 1 £ 10216
and a match tolerance of seven. The BESs44
from clones within
the fingerprint database were then matched to the tile path of human sequence clones.
Human tile path accessions and mouse BESs were repeat masked and then aligned by
BLASTN version 2.0a13MP-WashU, 10 June 1997 (http://blast.wustl.edu; W. Gish,
personal communication)45
. All matches longer than 100 bp, and with a significance value
above 0.01, were clustered using the initial fingerprint assembly. A ‘high score’ blast match
(that is, with a blast score .700) was required to anchor each mouse map contig to the
human tile path, supported by at least one other BLASTmatch (Fig. 1). Only matches with
unique locations in the human tile path were used. Matches between the mouse BESs and
human accessions had an average score of 627.29 and an average identity of 72.3%. Contigs
that were adjacent to each other on the basis of their alignment to the human sequence
were inspected manually and joined if fingerprints of the end clones overlapped with
probability scores of better than 1 £ 10212
. Markers were added to the map either by
electronic PCR31
, or by hybridization using overgo probes33
. Markers were incorporated
into the physical map only if they were uniquely placed within the map. Additional joins
were made between contigs that were adjacent according to consistent mouse marker data
if supported additionally by fingerprint overlaps with probability scores of better than
1 £ 10210
.
Coverage estimates
The exact length of 125 mouse clones, for which finished sequence was available (totalling
25,349 kb) was divided by the number of HindIII fragments (5,597) observed as single
bands in the fingerprints of these clones. This provided a conversion factor of 4.5 kb per
observed fragment, which was used to calculate all contig sizes in the consensus map as
listed in Table 2. An estimate of 2.8 Gb for the size of the mouse genome was obtained as
follows. Assembled mouse whole-genome shotgun data (November release, covering a
total of 2.41 Gb) contained 91% coverage of 41 Mb of finished mouse genomic sequence.
On this basis, the mouse genome is estimated to cover ,2.8 Gb (2.65 Gb euchromatin plus
heterochromatin). Length estimates for individual mouse chromosomes were taken from
refs 46 and 47, adjusted using the estimate of whole genome size described above (see Table
2). This result suggests that the mouse genome is shorter than the human genome.
Support for this was obtained by comparing contiguous and finished mouse and human
genomic sequence in 13 regions that could be aligned. The combined total lengths were
6.29 Mb (mouse) and 7.78 Mb (human) (ratio 0.8), although values for individual
segments varied between 0.73 and 1.06. These results are in line with previous
observations34,35,39,48–50
.
Received 15 April; accepted 4 July 2002; doi:10.1038/nature00957.
Published online 4 August 2002.
1. Fleischmann, R. D. et al. Whole-genome random sequencing and assembly of Haemophilus influenzae
Rd. Science 269, 496–512 (1995).
2. Churcher, C. et al. The nucleotide sequence of Saccharomyces cerevisiae chromosome IX. Nature 387,
84–87 (1997).
3. The yeast genome directory. Nature 387 (Suppl.), 5 (1997).
4. The C. elegans Sequencing Consortium Genome sequence of the nematode C. elegans: a platform for
investigating biology. Science 282, 2012–2018 (1998).
5. Adams, M. D. et al. The genome sequence of Drosophila melanogaster. Science 287, 2185–2195
(2000).
6. Lander, E. S. et al. Initial sequencing and analysis of the human genome. Nature 409, 860–921
(2001).
7. Venter, J. C. et al. The sequence of the human genome. Science 291, 1304–1351 (2001).
8. Collins, J. & Hohn, B. Cosmids: a type of plasmid gene-cloning vector that is packageable in vitro in
bacteriophage lambda heads. Proc. Natl Acad. Sci. USA 75, 4242–4246 (1978).
9. Shizuya, H. et al. Cloning and stable maintenance of 300-kilobase-pair fragments of human
DNA in Escherichia coli using an F-factor-based vector. Proc. Natl Acad. Sci. USA 89, 8794–8797
(1992).
10. Coulson, A., Sulston, J., Brenner, S. & Karn, J. Towards a physical map of the genome of the nematode
Caenorhabditis elegans. Proc. Natl Acad. Sci. USA 83, 7821–7825 (1986).
11. Olson, M. V. et al. Random-clone strategy for genomic restriction mapping in yeast. Proc. Natl Acad.
Sci. USA 83, 7826–7830 (1986).
12. Bentley, D. R. et al. The physical maps for sequencing human chromosomes 1, 6, 9, 10, 13, 20 and X.
Nature 409, 942–943 (2001).
13. McPherson, J. D. et al. A physical map of the human genome. Nature 409, 934–941 (2001).
14. Green, E. D. Strategies for the systematic sequencing of complex genomes. Nature Rev. Genet. 2,
573–583 (2001).
15. Dunham, I. et al. The DNA sequence of human chromosome 22. Nature 402, 489–495 (1999).
16. Waterston, R. & Sulston, J. E. The Human Genome Project: reaching the finish line. Science 282, 53–54
(1998).
17. Nadeau, J. H. & Taylor, B. A. Lengths of chromosomal segments conserved since divergence of man
and mouse. Proc. Natl Acad. Sci. USA 81, 814–818 (1984).
18. DeBry, R. W. & Seldin, M. F. Human/mouse homology relationships. Genomics 33, 337–351
(1996).
19. Nadeau, J. H. & Sankoff, D. Counting on comparative maps. Trends Genet. 14, 495–501 (1998).
20. Thomas, J. W. et al. Comparative genome mapping in the sequence-based era: early experience with
human chromosome 7. Genome Res. 10, 624–633 (2000).
21. Kim, J. et al. Homology-driven assembly of a sequence-ready mouse BAC contig map spanning
regions related to the 46-Mb gene-rich euchromatic segments of human chromosome 19. Genomics
74, 129–141 (2001).
22. Osoegawa, K. et al. Bacterial artificial chromosome libraries for mouse sequencing and functional
analysis. Genome Res. 10, 116–128 (2000).
23. Hardison, R. C., Oeltjen, J. & Miller, W. Long human–mouse sequence alignments reveal novel
regulatory elements: a reason to sequence the mouse genome. Genome Res. 7, 959–966 (1997).
24. Koop, B. F. et al. The human T-cell receptor TCRAC/TCRDC (Ca/Cd) region: organization, sequence,
and evolution of 97.6 kb of DNA. Genomics 19, 478–493 (1994).
25. Ashworth, L. K. et al. An integrated metric physical map of human chromosome 19. Nature Genet. 11,
422–427 (1995).
26. Kent, W. J. The BLAST-like alignment tool. Genome Res. 12, 656–664 (2002).
27. Copeland, N. G. et al. A genetic linkage map of the mouse: current applications and future prospects.
Science 262, 57–66 (1993).
28. Dietrich, W. F. et al. A comprehensive genetic map of the mouse genome. Nature 380, 149–152
(1996).
29. Cai, W. W. et al. An SSLP marker-anchored BAC framework map of the mouse genome. Nature Genet.
29, 133–134 (2001).
30. Hudson, T. J. et al. A radiation hybrid map of mouse genes. Nature Genet. 29, 201–205 (2001).
31. Schuler, G. D. Sequence mapping by electronic PCR. Genome Res. 7, 541–550 (1997).
32. Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. Basic local alignment search tool.
J. Mol. Biol. 215, 403–410 (1990).
33. Ross, M. T., LaBrie, S., McPherson, J. P. & Stanton, V. P. in Current Protocols in Human Genetics (eds
Dracopoli, N. C. et al.) 5.6.1–5.6.5 (Wiley, New York, 1999).
34. Puttagunta, R. et al. Comparative maps of human 19p13.3 and mouse chromosome 10 allow
identification of sequences at evolutionary breakpoints. Genome Res. 10, 1369–1380 (2000).
35. Carver, E. A. & Stubbs, L. Zooming in on the human–mouse comparative map: genome conservation
re-examined on a high-resolution scale. Genome Res. 7, 1123–1137 (1997).
36. Watkins-Chow, D. E. et al. Genetic mapping of 21 genes on mouse chromosome 11 reveals disruptions
in linkage conservation with human chromosome 5. Genomics 40, 114–122 (1997).
37. Stubbs, L. et al. Detailed comparative map of human chromosome 19q and related regions of the
mouse genome. Genomics 35, 499–508 (1996).
38. Dehal, P. et al. Human chromosome 19 and related regions in mouse: conservative and lineagespecific
evolution. Science 293, 104–111 (2001).
39. Pletcher, M. T., Wiltshire, T., Cabin, D. E., Villanueva, M. & Reeves, R. H. Use of comparative physical
and sequence mapping to annotate mouse chromosome 16 and human chromosome 21. Genomics 74,
45–54 (2001).
40. Oeltjen, J. C. et al. Large-scale comparative sequence analysis of the human and murine Bruton’s
tyrosine kinase loci reveals conserved regulatory domains. Genome Res. 7, 315–329 (1997).
41. Batzoglou, S., Pachter, L., Mesirov, J. P., Berger, B. & Lander, E. S. Human and mouse gene structure:
comparative analysis and application to exon prediction. Genome Res. 10, 950–958 (2000).
42. Marra, M. A. et al. High throughput fingerprint analysis of large-insert clones. Genome Res. 7,
1072–1084 (1997).
43. Soderlund, C., Humphray, S., Dunham, A. & French, L. Contigs built with fingerprints, markers, and
FPC V4. 7. Genome Res. 10, 1772–1787 (2000).
44. Zhao, S. et al. Mouse BAC ends quality assessment and sequence analyses. Genome Res. 11, 1736–1745
(2001).
45. Altschul, S. F. & Gish, W. Local alignment statistics. Methods Enzymol. 266, 460–480 (1996).
46. Nusbaum, C. et al. A YAC-based physical map of the mouse genome. Nature Genet. 22, 388–393
(1999).
47. Evans, E. P. in Genetic Variants of the Laboratory Mouse (eds Lyon, M. F., Rastan, S. & Brown, S. D. M.)
1446 (Oxford Univ. Press, New York, 1996).
48. Pletcher, M. T. et al. Chromosome evolution: the junction of mammalian chromosomes in the
formation of mouse chromosome 10. Genome Res. 10, 1463–1467 (2000).
49. Martindale, D. W. et al. Comparative genomic sequence analysis of the Williams syndrome region
(LIMK1-RFC2) of human chromosome 7q11.23. Mamm. Genome 11, 890–898 (2000).
50. DeSilva, U. et al. Generation and comparative analysis of approximately 3.3 Mb of mouse genomic
sequence orthologous to the region of human chromosome 7q11.23 implicated in Williams
syndrome. Genome Res. 12, 3–15 (2002).
articles
NATURE | VOL 418 | 15 AUGUST 2002 | www.nature.com/nature 749© 2002 NaturePublishing Group
Supplementary Information accompanies the paper on Nature’s website
(http://www.nature.com/nature). Updated views of the map are available from the authors’
websites (http://www.ensembl.org/Mus_musculus/cytoview and http://www.ncbi.nlm.nih.gov/
genome/guide/mouse), as is an archive version for this publication, plus comparisons and
discussion of genetic, RH and clone maps, and an interactive version of the synteny displays of
Fig. 5 (http://www.sanger.ac.uk/Projects/M_musculus/publications/fpcmap-2002).
Acknowledgements
The authors acknowledge the support of the Wellcome Trust, the National Institutes of
Health and the US Department of Energy. We are grateful to the web team at the Sanger
Institute for assistance with developing map displays, to P. Deloukas for RH map analysis,
and E. Arnold-Berkowits, S. Lo, J. Gill and all present and past members of the Institute for
Genomic Research BAC end sequencing team for the sequencing work.
Competing interests statement
The authors declare that they have no competing financial interests.
Correspondence and requests for materials should be addressed to D.R.B.
(e-mail: drb@sanger.ac.uk).
articles
NATURE | VOL 418 | 15 AUGUST 2002 | www.nature.com/nature750 © 2002 NaturePublishing Group
a Mouse–human homology map
b Human–mouse homology map
3 6
2
A
B
C
D
E
1
2
3
1
2
3
1
2
3
4
5
1
3
1
F
G
H
2
2
3
4
3
2
1
2
3
1
3
3
3
3
3
1
2
3
4
1
A
B
C
D
E
F
G
H
4
A
B
C
D
E
1
2
3
4
5
1
2
3
1
2
3
4
5
6
7
1
2
3
1
2
A
B
C
D
E
2
3
4
5
1
2
3
4
5
1
2
3
4
1
2
3
4
5
6
A
B
C
D
E
G
H
F
1
2
3
1
1
1
3
2
3
1
2
3
A
B
C
D
E
F
G
5 6
A
B
C
D
F
G
2
3
1
1
1
3
2
2
3
5
1
2
3
E
1
2
3
A
7
B
C
D
E
F
2
1
4
1
3
2
5
31
1
2
3
2
1
3
4
A
21 4 5 7 8
p
q
q
q
q
q
q
q
q
p
p
p
p
p
p
p
© 2002 NaturePublishing Group
9 14 16
10
2
2
3
1
1
4
1
35
1
3
2
3
4
2
2
1
3
2
A
B
C
D
11
A
B
C
D
E
1
3
5
1
2
3
1
5
4
2
2
4
12
B
C
E
F
2
3
1
1
1
3
2
3
1
2
A
D
13
2
5
3
1
2
1
4
2
3
1
2
A
B
C
D
14
A
B
C
D
E
2
3
1
3
1
1
1
1
3
4
2
2
3
2
A
B
C
D
E
F
8
A
B
C
D
E
3
1
1
3
2
1
2
3
1
2
2
3
4
1
2
9
B
C
E
F
2
4
2
1
2
3
3
4
1
2
3
5
1
4
A
D
10 11 12 13 15
q
p
q
q
q
p
p
p
p
p
q
q
q
q
p p
© 2002 NaturePublishing Group
17 18 19 212120 22
Y
A
B
C
D
15
A
B
C
D
E
F
2
3
1
1
2
3
1
2
3
16
A
B
C
5
1
1
1
2
4
5
4
3
4
17
A
B
C
D
E
2
1
1
3
2
3
4
5
19
A
B
C
D
1
1
3
2
18
A
B
C
D
E
2
3
1
1
1
2
3
4
X
A
B
C
D
E
F
Xp11.1
Xp11.2
Xp11.3/.4
Xq23
Xq28
Xp21
cen
Xq22
Xp11.3
Xp22.33
Xp22.1
Xp22.3
X Y
q
p
p
p
q
q
q
q
p
© 2002 NaturePublishing Group