PDF - Martin Krzywinski

1

VIZBI 2013

CANADA’S MICHAEL SMITH GENOME SCIENCES CENTER

BC CANCER RESEARCH CENTER

MARTIN KRZYWINSKI

visual design principles

C O M M U N I C A T I N G C L E A R L Y

VIZBI 2013

This is an extended versionof my VIZBI 2013 keynote presentation.

vizbi.org/2013

Additional redesign examples are included.

VIZBI 2013

Do not rely solely on your personal aesthetic.

Strive for simplicity and clarity.

YOU WANT TO BE A BETTER COMMUNICATOR

NOT A BETTER ARTIST

VIZBI 2013

It’s not all about taste.

OBJECTIVE ASPECTS OF DESIGN

VIZBI 2013

VIZBI 2013

TAGLINE

VIZBI 2013

VIZBI 2013

Excellent organization and consistency. Vertical lines cue continuity. Good use of color.

Samollow, P.B., The opossum genome: insights and opportunities from an alternative mammal. Genome Res, 2008. 18(8): p. 1199-215.

Chartjunk plentiful. Screaming ornamental and redundant elements. Text inconsistent and illegible.

Gentles, A.J., et al., Evolutionary dynamics of transposable elements in the short-tailed opossum Monodelphis domestica. Genome Res, 2007. 17(7): p. 992-1004.

VIZBI 2013

In either, we don’t always know the end product.

But we must understand how we might get there.

SCIENCE IS A PROCESS

DESIGN IS A PROCESS

VIZBI 2013

you have data

you may have a message

you need to create a !gure (or think you do)

VIZBI 2013

you select your encoding

e.g. sequence logo

VIZBI 2013

you generate the image with an application that supports the encoding

VIZBI 2013

you write a legend, making upfor things that are not obvious

William Ray, http://www.biovis.net/year/2013/info/redesign-contest

Sequence logos showing the amino acid usage in the adenylate kinase lid (AKL) domain. (A) Across all organisms. (B) from Gram-negative bacteria. (C) from Gram-positive bacteria. The ADK lid domain structure is universally conserved, but is stabilized in the Gram-negatives by a hydrogen bonding network between residues 4, 7, 9, 24, 27, and 29 (and several other residues in some organisms), while the Gram-positives are stabilized by a bound metal ion, tetrahedrally coordinated by the Cysteines at 4, 7, 24 and 27. The identities of several other positions (eg 5, 8, 30, 32) are differentially constrained in each subfamily as well, apparently due to steric requirements of the stabilizing residues.

VIZBI 2013

your !gure is done

how do you knowwhether the creation process

was successful?

VIZBI 2013

The desire for a !gure is not always proportional to its utility.

CREATE VISUALS WHEN NECESSARY

VIZBI 2013

TAGLINE

VIZBI 2013

TAGLINE

lushlush.livejournal.com/190093.html

via @jayjacobs, https://twitter.com/jayjacobs

IS ABSOLUTE ACCURACY ALWAYS IMPORTANT?

https://twitter.com/jayjacobs

https://twitter.com/jayjacobs

accuracy does not always add to utility

NO

VIZBI 2013

In the age of big data, !gures should be worth more than 1,000 words.

INFORMATIVE + INFORMATION-RICH

VIZBI 2013

strive to give your viewer the greatest number of useful ideas

in the shortest time with the least ink

in the smallest space

Tufte, E. The Visual Display of Quantitative Information (Graphic Press, Cheshire, Connecticut, USA, 2007).

INFORMATION-RICH INFORMATIVE

CONFUSING GRATUITOUSEFFECTIVE

VISUALIZATION SWEET SPOT

CLEAR MESSAGE

HIGH DATA-TO-INK RATIO

ACCESSIBLE COMPLEXITY

Alekseyev, M.A. and P.A. Pevzner, Breakpoint graphs and ancestral genome reconstructions. Genome Res, 2009. 19(5): p. 943-57.

INFORMATION-RICH, NOT INFORMATIVE

Harr, B., Genomic islands of differentiation between house mouse subspecies. Genome Res, 2006. 16(6): p. 730-7.

INFORMATIVE, NOT INFORMATION-RICH

Hurle, B., W. Swanson, and E.D. Green, Comparative sequence analyses reveal rapid and divergent evolutionary changes of the WFDC locus in the primate lineage. Genome Res, 2007. 17(3): p. 276-86.

INFORMATION-RICH AND INFORMATIVE

(left) Synastry chart. http://sasstrology.com/2011/03/the-astrology-of-marriage-in-the-royal-family-a-suitable-girl-and-the-bit-on-the-side.html(right) Shakhnovich, B.E. and E.V. Koonin, Origins and impact of constraints in evolution of gene families. Genome Res, 2006. 16(12): p. 1529-36.

JUNK SCIENCE REAL SCIENCE

VIZBI 2013

don’t merely display data

explain it

VIZBI 2013

Stick to it.

KNOW YOUR MESSAGE

VIZBI 2013

TAGLINE

Kawai, J. and Y. Hayashizaki, DNA book. Genome Res, 2003. 13(6B): p. 1488-95.

Kawai, J. and Y. Hayashizaki, DNA book. Genome Res, 2003. 13(6B): p. 1488-95.Fig 1. redesigned by M Krzywinski

RAW DATA

DISCRETIZED

BINNED

TREND

NO CLEAR MESSAGE.UNKNOWN. READER ISON THEIR OWN.

THREE RANGES ARE IMPORTANT.INDIVIDUAL VALUES WITHIN ARANGE ARE NOT.

THERE ARE FEWER MEDIUM-SIZED VALUES.

THERE ARE SIGNIFICANTLYFEWER MEDIUM-SIZED VALUES.

SCALE

DISTRIBUTION

SIGNIFICANCE

WHAT IS SHOWN? WHAT IS COMMUNICATED? WHAT IS INTERPRETED?

30 60

EXPECTED

*

30 60

12 54 82 29 25 22 67 61 23 79

0-30

31-60

61-100

ACHIEVE FOCUS BY AGGREGATING

CONTEXT MUST NEVER DILUTE MESSAGE

no message message in isolationmessage in context

GENE DISEASE GENE COMPLEX

CONNECTING GENE

SYSTEM GENE COMPLEX

GENES INCOMPLEX

CONNECTINGGENE

S1

S1

S2

S3

S4

A

AG1

G1

SYSTEM

G2 G3

G5 G6 G7

G8

G4

G2G3

G3

G8

G4

G5

G6G7

CC

D

D

B

B

B

F

FE

ES4

S3

S3

S2

VIZBI 2013

Revise and redraw.

STRIVE FOR CLEAR COMMUNICATION

VIZBI 2013

to explore data, choose effective encoding

to communicate concepts, use effective design

Tufte, E. The Visual Display of Quantitative Information (Graphic Press, Cheshire, Connecticut, USA, 2007).

HOW TO APPROACH VISUALIZATION

show the data

induce viewer to think about substance rather than methodology

encourage eye to compare different pieces of data

avoid distorting what the data represents

present many numbers in a small space

make large data sets coherent

reveal data at several levels of detail – broad overview and !ne structure

TOP-DOWN

redundancyconsistencyconciseness

clarityfocus & emphasis

salience & relevancetruth, accuracy & detail

BOTTOM-UP

data encodingsymbolscolortypefacearrowsline weightalignment

I N FORMS

SAT I S F I E S

VIZBI 2013

Correctness does not ensure comprehension.

ACCURACY ISN’T EVERYTHING

VIZBI 2013

Buffalo buffalo Buffalo buffalo buffalo buffalo Buffalo buffalo

http://en.wikipedia.org/wiki/Buffalo_buffalo_Buffalo_buffalo_buffalo_buffalo_Buffalo_buffalo



VIZBI 2013

New York bison whom other New York bison bully, themselves bully New York bison.




Alekseyev, M.A. and P.A. Pevzner, Breakpoint graphs and ancestral genome reconstructions. Genome Res, 2009. 19(5): p. 943-57.

BUFFALO BUFFALO OF VISUALIZATION

VIZBI 2013

Be aware of bias in evaluating effectiveness of visual forms.

SATISFY YOUR AUDIENCE, NOT YOURSELF.

Elting et al. BMJ 316 (1999).

Good prognosis 30 35 11

Poor prognosis 45 25 12

Total 38

30

20

50 60 12

Investigational treatmentConventional treatment

Total no % Fail Total no % Fail

Good prognosis Good prognosis

Investigational treatment(n = 50)

Conventional treatment(n = 60)

Poor prognosis

Response Fail

Poor prognosis

Investigational treatmentConventional treatment

Response Fail

No

of p

atie

nts

0

20

30

40

10

Goodprognosis

Poorprognosis

Poorprognosis

Goodprognosis

Response Fail

Good prognosis Poor prognosis

Investigational treatment

Conventional treatment

bar graph

table pie chart

icon graph


%

68

82

56

56

30 30 35 1120 45 25 1250 38 60 12

accuracy preference

62

24

14

0

%

VIZBI 2013

“...eight voicedconsiderable contemptfor the [icon] display.”


VIZBI 2013

“... icon displays were oftenpreferred by nurses, students,

... but were consideredunacceptable by physicians.”


Shephard’s tables, http://www.michaelbach.de/ot/sze_shepardTables/

WE ARE EASILY DECEIVED

DO NOT BE CHARMED BY INEFFECTIVE FORMS

Krzywinski et al. Brie!ngs in Bioinformatics (2011).

THE INCOMPARABLE NETWORK

VIZBI 2013

RESPECT HUMAN VISUAL LIMITATIONS

VIZBI 2013

TAGLINE

VIZBI 2013

REDUNDANCY

VIZBI 2013

TAGLINE

http://elenamiska.com/AIDS-Poster

VIZBI 2013

Utility of RNA Sequencing for Analysis of Maize Reproductive Transcriptomes Plant Gen. 2011 4: 191–203 10.3835/plantgenome2011.05.0015

VIZBI 2013

Peri S, Navarro JD, Amanchy R, Kristiansen TZ, Jonnalagadda CK, et al. (2003) Development of human protein reference database as an initial platform for approaching systems biology in humans. Genome Res 13: 2363-2371.

EGF

EGFR

GAP

RAS RHOGAPEPHB2

EGF

EGFR

GAP

RAS RHOGAPEPHB2 RHOGAPEPHB2

2.9 10.0

10.2 10.1

14.6

106 107 174!LRGB

158 179 232

176 229 158

232 163 163 180 180 180

107 185 103

248 54 81

5.8

3.8

poor contrast improved contrast improved contrastand consistency

VIZBI 2013

UNIFORM PALETTE

UNIFORM SPACING BREWER SPACING

9 TONE GREY BREWER PALETTE

0

0

255128 191 223159966432

2552402171891501158237

grey level

poortoneseparation

large tonedi!erence

EGF

EGFR

GAP

RAF

CREB

RAS RAS

PKC

CKII

EGF

EGFR

GAP

RAF

CREB

PKC

CKII

VIZBI 2013

Patterns are hard to see when variation is due to both data and formatting.

DATA INFORMS VARIATION

HOW MANY DOTS?

AVOID EXCESS DEGREES OF FREEDOM

Sharov AA, Dudekula DB, Ko MS (2005) Genome-wide assembly and analysis of alternative transcripts in mouse. Genome Res 15: 748-754.

VIZBI 2013

Avoid the use of similar encodings for independent variables.

CONSISTENCY

VIZBI 2013

TAGLINE

Lewis EK, Haaland WC, Nguyen F, Heller DA, Allen MJ, et al. (2005) Color-blind "uorescence detection for four-color DNA sequencing. Proc Natl Acad Sci U S A 102: 5346-5351.

VIZBI 2013

Lewis EK, Haaland WC, Nguyen F, Heller DA, Allen MJ, et al. (2005) Color-blind "uorescence detection for four-color DNA sequencing. Proc Natl Acad Sci U S A 102: 5346-5351.

AF405

CY 5.5

BODIPY-FL

6-ROX

685 nm

399 nm

488 nm

594 nm

685 nm

399 nm

488 nm

594 nm

C

A

G

T

BODIPY-FL

6-ROX

CY 5.5

AF405

disambiguatedhomonymous encoding

VIZBI 2013

TAGLINE

McEwen GK, Woolfe A, Goode D, Vavouri T, Callaway H, et al. (2006) Ancient duplicated conserved noncoding elements in vertebrates: a genomic and functional analysis. Genome Res 16: 451-465.

VIZBI 2013

McEwen GK, Woolfe A, Goode D, Vavouri T, Callaway H, et al. (2006) Ancient duplicated conserved noncoding elements in vertebrates: a genomic and functional analysis. Genome Res 16: 451-465.

arrows too complex simpli!ed arrows

Mya

EVOLUTION RATE

species 1 species 2

ancestor

early ancestorA

C1

C2

D1

D2

B

CNE EVOLUTION

A

B

EVOLUTION RATECNE EVOLUTION

rapid

slow

C1

C2

D2

D2species 1 species 2

early ancestor

ancestor

650 Mya

450 Mya

650

450

VIZBI 2013

Don’t repeat yourself. Don’t repeat yourself.

REDUNDANCY

Zinc !nger exon analysis for ZNF493 and ZNF738, two divergent genes from the ZNF431 clade. Hamilton, A.T., et al., Evolutionary expansion and divergence in the ZNF91 subfamily of primate-speci!c zinc !nger genes. Genome Res, 2006. 16(5): p. 584-94.

Nikaido M, Rooney AP, Okada N (1999) Phylogenetic relationships among cetartiodactyls based on insertions of short and long interspersed elements: Hippopotamuses are the closest extant relatives of whales. Proceedings of the National Academy of Sciences 96: 10261-10266.

: ARE

: ARE 2

: BovB LINE

: LINE 1

ARE

ARE 2

LINE 1

BovB LINE

overwhelming simpli!ed

ARE

ARE 2

LINE 1

BovB LINE

integrated key

VIZBI 2013

Use the fewest elements possible—keep data-to-ink ratio high

CONCISENESS

Sharov AA, Dudekula DB, Ko MS (2005) Genome-wide assembly and analysis of alternative transcripts in mouse. Genome Res 15: 748-754.

Sharov AA, Dudekula DB, Ko MS (2005) Genome-wide assembly and analysis of alternative transcripts in mouse. Genome Res 15: 748-754.redesign by M Krzywinski

A mix-and-match model for prokaryotic genome evolution. Charlebois, R.L. and W.F. Doolittle, Computing prokaryotic gene ubiquity: rescuing the core from extinction. Genome Res, 2004. 14(12): p. 2469-77.

A mix-and-match model for prokaryotic genome evolution. Charlebois, R.L. and W.F. Doolittle, Computing prokaryotic gene ubiquity: rescuing the core from extinction. Genome Res, 2004. 14(12): p. 2469-77.redesign by M Krzywinski

metabolism

envelope

translation

simple visual vocabulary

MIX!AND!MATCH MODELFUNCTIONAL GENES

analogoushomologous

loss of function

VIZBI 2013

Use the fewest elements possible—keep data-to-ink ratio high.

Shelter your reader from unnecessary complexity.

REMOVE TO IMPROVE

Telomere sequence gaps. Riethman, H., et al., Mapping and initial analysis of human subtelomeric sequence assemblies. Genome Res, 2004. 14(1): p. 18-28.

Telomere sequence gaps. Riethman, H., et al., Mapping and initial analysis of human subtelomeric sequence assemblies. Genome Res, 2004. 14(1): p. 18-28.redesign by M Krzywinski

Resequencing with naïve and log pool designs. Prabhu, S. and I. Pe'er, Overlapping pools for high-throughput targeted resequencing. Genome Res, 2009. 19(7): p. 1254-61.

Resequencing with naïve and log pool designs. Prabhu, S. and I. Pe'er, Overlapping pools for high-throughput targeted resequencing. Genome Res, 2009. 19(7): p. 1254-61.redesign by M Krzywinski

DISJOINT OVERLAPPINGbusy

improved

POOLS

INDIVIDUALS INDIVIDUALS

10000000

10000000

01000000

01000000

00100000

00100000

00010000

00010000

00001000

00001000

00000100

00000100

00000010

00000010

00000001

00000001

POOLS

00001111

00011110

00101101

00111100

01001011

01011010

01101001

01111000

10000111

10010110

10100101

10110100

11000011

11000010

11010010

11110000

Valouev A, Ichikawa J, Tonthat T, Stuart J, Ranade S, et al. (2008) A high-resolution, nucleosome position map of C. elegans reveals a lack of universal sequence-dictated positioning. Genome Res 18: 1051-1063.

detail exposedPositioned and

Uniformly SpacedNot Positioned but Uniformly Spaced

Not Positioned and Not Uniformly Spaced

Positioned but Not Uniformly Spaced

detail encapsulatedPOSITIONED

UNIFORMLYSPACED

Y

Y

N

N

VIZBI 2013

Make sure that elements are visible and unobscured.

Don’t count on your audience to !gure out what you mean. Say it.

CLARITY

VIZBI 2013

Nadeau JH, Burrage LC, Restivo J, Pao YH, Churchill G, et al. (2003) Pleiotropy, homeostasis, and functional networks based on assays of cardiovascular traits in genetically randomized populations. Genome Res 13: 2082-2091.

concentrichypertrophy

eccentrichypertrophy

normal

normalconcentrichypertrophy


normalconcentrichypertrophy


perceptually challenging helpful, but not explicit helpful and explicit

VIZBI 2013

Match the pertinence of an object with its visual salience.

Apply visual organization Gestalt principles.

FOCUS & EMPHASIS

Marcin Ryczek, http://plfoto.com/182312

Marcin Ryczek, http://plfoto.com/182312

Bechah, Y., et al., Genomic, proteomic, and transcriptomic analysis of virulent and avirulent Rickettsia prowazekii reveals its adaptive mutation capabilities. Genome Res, 2010. 20(5): p. 655-63.

EVERYTHING IS EMPHASIZED

Severino, S., Messina, R. (2011). Group Dynamics in On-Line and Face-to-Face Interactions: An Experimental Study on Learning Methods. Sociology Mind. Vol.1, No.2, 65-73 DOI:10.4236/sm.2011.12008

NOTHING IS EMPHASIZED

Fecteau JH, Munoz DP (2006) Salience, relevance, and !ring: a priority map for target selection. Trends Cogn Sci 10: 382-390.Yantis S (2005) How visual salience wins the battle for awareness. Nat Neurosci 8: 975-977.

MATCH VISUAL SALIENCE TO RELEVANCE

ATTENTIONAL CAPTURE

inhibitory interactionlow high

neural responselow high

DISTINCT

!t

HETEROGENEOUS

!t

HIGHSALIENCE

LOW

Vettore AL, da Silva FR, Kemper EL, Souza GM, da Silva AM, et al. (2003) Analysis and functional annotation of an expressed sequence tag collection for tropical crop sugarcane. Genome Res 13: 2725-2735.Bono H, Yagi K, Kasukawa T, Nikaido I, Tominaga N, et al. (2003) Systematic expression pro!ling of the mouse transcriptome using RIKEN cDNA microarrays. Genome Res 13: 1318-1323.Tenney AE, Wu JQ, Langton L, Klueh P, Quatrano R, et al. (2007) A tale of two templates: automatically resolving double traces has many applications, including ef!cient PCR-based elucidation of alternative splices. Genome Res 17: 212-218.

USE BREWER PALETTES

one colordominates

di!cult todistinguish murky

A B C D E

A B C D E

screen blend mode3 4 5 5 6 3

set21 2 3 4 5 6 7 8

recolored with Brewer palettes

VIZBI 2013

Use symbols that intuitively encode related concepts.

HIERARCHY AND PRIORITY

Zheng, D., et al., Pseudogenes in the ENCODE regions: consensus annotation, analysis of transcription, and evolution. Genome Res, 2007. 17(6): p. 839-51.



Nelander S, Mostad P, Lindahl P (2003) Prediction of cell type-speci!c gene modules: identi!cation and initial characterization of a core set of smooth muscle-speci!c genes. Genome Res 13:1838-1854.

all genes

speci!c in situ result

selective in situ result

non-selective in situ result

speci!c in literature

non-speci!c in literature

no hierarchy

speci!c

selective

non-selective

non-speci!c

gene

in litin situ

hierarchical layout

speci!c

selective

non-selective

non-speci!c

gene

in litin situ

hierarchical layout and encoding

VIZBI 2013

VISUAL ABSTRACT REDESIGN

JH Barlow et al (2013) Identi!cation of Early Replicating Fragile Sites that Contribute to Genome Instability. Cell 152:620-632.

JH Barlow et al (2013) Identi!cation of Early Replicating Fragile Sites that Contribute to Genome Instability. Cell 152:620-632.

JH Barlow et al (2013) Identi!cation of Early Replicating Fragile Sites that Contribute to Genome Instability. Cell 152:620-632.redesign by M Krzywinski








VIZBI 2013

Identifying a deliverable message.

VISUAL ABSTRACT CREATION

Chun HJE, Khattra J, Krzywinski M, Aparicio SA, Marra MA (Dellaire GD, Berman JN, Arceci RJ. editors). Cancer Genomics. 1st ed. Elsevier; 2013.

KEYWORDS

cancercancer genomicstumor heterogeneitynext-generation sequencingsecond-generation sequencingthird-generation sequencingmutation discoverywhole genome sequencingsingle molecule sequencingsingle cell sequencingpersonalized medicine

http://www.elsevier.com/authors/graphical-abstract

ELSEVIER FIGURE RESTRICTION

Please note that your image will be scaled proportionally to !t in the available window on ScienceDirect, a 500 by 200 pixel rectangle.



VIZBI 2013



vizbi.org/2013/Program

Yates LR, Campbell PJ. Evolution of the cancer genome. Nat Rev Genet 2012 Nov;13(11):795-806.

VIZBI 2013




Ding L, Ley TJ, Larson DE, Miller CA, Koboldt DC, Welch JS, et al. Clonal evolution in relapsed acute myeloid leukaemia revealed by whole-genome sequencing. Nature. 2012 Jan 26;481(7382):506–10.

VIZBI 2013




Chun HJE, Khattra J, Krzywinski M, Aparicio SA, Marra MA (Dellaire GD, Berman JN, Arceci RJ. editors). Cancer Genomics. 1st ed. Elsevier; 2013.

VIZBI 2013




Chun HJE, Khattra J, Krzywinski M, Aparicio SA, Marra MA (Dellaire GD, Berman JN, Arceci RJ. editors). Cancer Genomics. 1st ed. Elsevier; 2013.Early version of visual abstract.

VIZBI 2013





VIZBI 2013





VIZBI 2013





VIZBI 2013





VIZBI 2013





VIZBI 2013





VIZBI 2013




Chun HJE, Khattra J, Krzywinski M, Aparicio SA, Marra MA (Dellaire GD, Berman JN, Arceci RJ. editors). Cancer Genomics. 1st ed. Elsevier; 2013.Final version of visual abstract.

VIZBI 2013

Chun HJE, Khattra J, Krzywinski M, Aparicio SA, Marra MA (Dellaire GD, Berman JN, Arceci RJ. editors). Cancer Genomics. 1st ed. Elsevier; 2013.Early (top) vs !nal (bottom) versions of visual abstract.

VIZBI 2013

Rational visual vocabulary for cell types.

REDESIGN—IN HOUSE

Chun HJE, Khattra J, Krzywinski M, Aparicio SA, Marra MA (Dellaire GD, Berman JN, Arceci RJ. editors). Cancer Genomics. 1st ed. Elsevier; 2013.Early version of Fig 1.






VIZBI 2013

Chun HJE, Khattra J, Krzywinski M, Aparicio SA, Marra MA (Dellaire GD, Berman JN, Arceci RJ. editors). Cancer Genomics. 1st ed. Elsevier; 2013.Early (left) vs !nal (right) version of Fig 1 legend.

VIZBI 2013

Assume your audience is smart but impatient.

REDESIGN—FROM LITERATURE

Recurrent R-spondin fusions in colon cancer. Somasekar Seshagiri et al. Nature 488, 660–664 (30 August 2012) doi:10.1038/nature11282

http://www.nature.com/nature/journal/v488/n7413/full/nature11282.html


http://www.nature.com/nature/journal/v488/n7413/fig_tab/nature11282_F3.html#auth-1








5 o

5 o

9 o

13o

10o

41o43o

7 o

12 o

19o

11 o

36o




















what is the core message? structure and evidence of a gene fusion

what is important? gene name and orientation location of breakpoint change in orientation, if any local sequence context supporting evidence

what is not important, or peripheral? gene size gene location gene model (learn to let go)

Recurrent R-spondin fusions in colon cancer. Somasekar Seshagiri et al. Nature 488, 660–664 (30 August 2012) doi:10.1038/nature11282Redesign by M Krzywinski

6q22.3

REFERENCE

SAMPLE

TGCATCCTAACGTTAGTCAAGGCTGCCAAGGAGGCTGTGCAACATGCTCATCTCCTGGGATCGGCCCAAGGCCAGTTCTCCGCAG






6q22.3

REFERENCE

SAMPLE


RSPO3 PTPRK802kb






6q22.3

REFERENCE

SAMPLE


RSPO3 PTPRK802kb

1 2 3 3 2 1






6q22.3

REFERENCE

SAMPLE


RSPO3 PTPRK802kb

2 31

1 2 3 3 2 1






6q22.3

REFERENCE

SAMPLE


RSPO3 PTPRK802kb

2 31

1 2 3 3 2 1






6q22.3

REFERENCE

SAMPLE


RSPO3 PTPRK802kb

2 31

1 2 3 3 2 1






6q22.3

REFERENCE

SAMPLE


RSPO3 PTPRK802kb

2 31

1 2 3 3 2 1





144

PDF - Martin Krzywinski

Documents