User:Was a bee/Gene

1. Test

A test trying to put a marker icon automatically onto the chromosome ideogram image to show location of gene (based on basepair position data stored in wikidata).

In other words, trying to make the image like below automatically.

Currently in Wikimedia Commons, there are about 100 ideogram images which are used to show gene position. See commons:category:Human chromosome ideograms which indicates gene location.

This is marker icon

This is plain ideogram of chromosome 7. Whole ideogram set is here commons:Template:Human chromosome ideograms in svg.

2. Result

Test result

Bp start	155,799,986
Bp end	155,812,273

Good (see the test case box at the right)

3. Calculation detail

The position where the marker should be put is calculated as follows. Although math expression looks something complex, actual calculation is not so complex. The concepts which are used here are basically only plus and minus for calculating length, and multiplication and division for calculating scaling.

Calculation detail

Other concepts used here are... Dividing sum of gene-start and gene-end by 2 is to get mid point of the gene. Dividing arrow-width by 2 is because image position is set by leftist point of the image, not by the center of the image. Conditional branch for arrow-width is just a technical problem which is needed to choose different shape of rectangles among these (commons:Template:Red rectangle series) as a marker depending on target gene length.

Calculation algorithm is as follows.... $PositionSetX={\Bigg \{}pterPx+{\bigg [}(qterPx-pterPx)\times {\frac {\frac {({\color {Red}GeneStart}+{\color {Red}GeneEnd})}{2}}{\color {Red}ChromosomeLength(n)}}{\bigg ]}{\Bigg \}}\times {\frac {ImageShownWidth}{ImageOriginalWidth}}-{\frac {\color {Blue}ArrowWidth}{2}}$

${\color {Blue}ArrowWidth^{*}}=(qterPx-pterPx)\times {\frac {({\color {Red}GeneEnd}-{\color {Red}GeneStart})}{\color {Red}ChromosomeLength(n)}}\times {\frac {ImageShownWidth}{ImageOriginalWidth}}$

${\color {Blue}ArrowWidth}={\begin{cases}2&\ ({\color {Blue}ArrowWidth^{*}}<2)\\\lceil {\color {Blue}ArrowWidth^{*}}\rceil &\ (otherwise.\lceil \;\rceil \;is\;ceiling\;function)\end{cases}}$

Where red terms are variables retrieved from Wikidata, the blue term is the term calculated based on variables retrieved from Wikidata, and other black terms are constants.

${\color {Red}GeneStart}$ : Gene start position from the terminus of p-arm. (unit: basepair, example: 155799986 from wikidata:Q14860072)
${\color {Red}GeneEnd}$ : Gene end position from the terminus of p-arm. (unit: basepair, example: 155812273 from wikidata:Q14860072)
${\color {Red}ChromosomeLength(n)}$ : Length of the $n$ th chromosome which contains the target gene. (unit: basepair, example: 159345973 for chromosome 7)
$pterPx$ : Horizontal position of pter (tip of p arm/short arm) in ideogram image (unit: pixel, example: 6 px)
$qterPx$ : Horizontal position of qter (tip of q arm/long arm) in ideogram image (unit: pixel, example: 1109 px)
$ImageOriginalWidth$ : Width of ideogram image (unit: pixel, example: 1125 px)

$ImageShownWidth$ : Shown width of ideogram image in Wikipedia page (unit: pixel, example: 300 px)
${\color {Blue}ArrowWidth^{*}}$ : Calculated marker width proportional to the gene length. This is generally too small (e.g. 0.05px) and non integer. (unit: pixel)
${\color {Blue}ArrowWidth}$ : Actual marker width shown in Wikipedia page. Minimum value is 2 and always integer (using ceiling function). To say, 2, 3, 4, 5...(unit: pixel)

By substituting the terms with example values, we get...

$PositionSetX={\Bigg \{}6+{\bigg [}(1109-6)\times {\frac {\frac {({\color {Red}GeneStart}+{\color {Red}GeneEnd})}{2}}{\color {Red}ChromosomeLength(n)}}{\bigg ]}{\Bigg \}}\times {\frac {300}{1125}}-{\frac {\color {Blue}ArrowWidth}{2}}$

$\qquad \qquad \qquad \ ={\Bigg \{}6+{\bigg [}{\frac {1103}{2}}\times {\frac {{\color {Red}GeneStart}+{\color {Red}GeneEnd}}{\color {Red}ChromosomeLength(n)}}{\bigg ]}{\Bigg \}}\times {\frac {300}{1125}}-{\frac {\color {Blue}ArrowWidth}{2}}$

$\qquad \qquad \qquad \ ={\bigg (}6\times {\frac {300}{1125}}{\bigg )}+{\bigg (}{\frac {1103}{2}}\times {\frac {300}{1125}}\times {\frac {{\color {Red}GeneStart}+{\color {Red}GeneEnd}}{\color {Red}ChromosomeLength(n)}}{\bigg )}-{\frac {\color {Blue}ArrowWidth}{2}}$

$\qquad \qquad \qquad \ ={\bigg (}{\frac {1103\times 300}{2\times 1125}}\times {\frac {{\color {Red}GeneStart}+{\color {Red}GeneEnd}}{\color {Red}ChromosomeLength(n)}}{\bigg )}+{\bigg (}{\frac {6\times 300}{1125}}{\bigg )}-{\frac {\color {Blue}ArrowWidth}{2}}$

$\qquad \qquad \qquad \ ={\bigg (}{\frac {330900}{2250}}\times {\frac {{\color {Red}GeneStart}+{\color {Red}GeneEnd}}{\color {Red}ChromosomeLength(n)}}{\bigg )}+{\bigg (}{\frac {1800}{1125}}{\bigg )}-{\frac {\color {Blue}ArrowWidth}{2}}$

$\qquad \qquad \qquad \ ={\bigg (}147.0666...\times {\frac {{\color {Red}GeneStart}+{\color {Red}GeneEnd}}{\color {Red}ChromosomeLength(n)}}{\bigg )}+1.6-{\frac {\color {Blue}ArrowWidth}{2}}$

$\qquad \qquad \qquad \ ={\bigg (}{\frac {{\color {Red}GeneStart}+{\color {Red}GeneEnd}}{\color {Red}ChromosomeLength(n)}}\times 147.0666...{\bigg )}+1.6-{\frac {\color {Blue}ArrowWidth}{2}}$

${\color {Blue}ArrowWidth^{*}}=(1109-6)\times {\frac {({\color {Red}GeneEnd}-{\color {Red}GeneStart})}{\color {Red}ChromosomeLength(n)}}\times {\frac {300}{1125}}$

$\qquad \qquad \quad \;\;\;\;={\frac {({\color {Red}GeneEnd}-{\color {Red}GeneStart})}{\color {Red}ChromosomeLength(n)}}\times {\frac {1103\times 300}{1125}}$

$\qquad \qquad \quad \;\;\;\;={\frac {({\color {Red}GeneEnd}-{\color {Red}GeneStart})}{\color {Red}ChromosomeLength(n)}}\times {\frac {330900}{1125}}$

$\qquad \qquad \quad \;\;\;\;={\frac {({\color {Red}GeneEnd}-{\color {Red}GeneStart})}{\color {Red}ChromosomeLength(n)}}\times 294.133..$

The form in the last lines of each equations are used in program.

Actual calculation is like as follows for SHH gene (wikidata:Q14860072).

Firstly we need to calculate arrow-width.

${\color {Blue}ArrowWidth^{*}}={\frac {({\color {Red}155812273}-{\color {Red}155799986})}{\color {Red}159345973}}\times 294.133..$

$\qquad \qquad \quad \;\;\;\;={\frac {{\color {Red}12287}\times 294.133..}{\color {Red}159345973}}$

$\qquad \qquad \quad \;\;\;\;={\frac {\color {Red}3614012.171}{\color {Red}159345973}}$

$\qquad \qquad \quad \;\;\;\;={\color {Red}0.022}$

Here we got 0.022px for marker image width for SHH gene. Is this wrong? No. This result is from tha fact that most genes are very short compared to whole chromosome length. If whole chromosome is shown in about 300px, most human genes (≒10kb) span only from 0.01px to 0.05px, depending on whole chromosome length. So third equation do job here. $({\color {Blue}ArrowWidth^{*}}<2)$ , to say $({\color {Red}0.022}<2)$ , is true. Hence we get....

${\color {Blue}ArrowWidth}={\color {Red}2}$

Arrow-width is 2px. Then we can calculate arrow position using this value.

$PositionSetX={\bigg (}{\frac {{\color {Red}155799986}+{\color {Red}155812273}}{\color {Red}159345973}}\times 147.0666...{\bigg )}+1.6-{\frac {\color {Red}2}{2}}$

$\qquad \qquad \qquad \ ={\bigg (}{\frac {\color {Red}311612259}{\color {Red}159345973}}\times 147.0666...{\bigg )}+1.6-{\color {Red}1}$

$\qquad \qquad \qquad \ ={\bigg (}{\color {Red}1.95557034253}\times 147.0666...{\bigg )}+{\color {Red}0.6}$

$\qquad \qquad \qquad \ ={\color {Red}287.599081336}+{\color {Red}0.6}$

$\qquad \qquad \qquad \ ={\color {Red}288.199081336}$

$\qquad \qquad \qquad \ \fallingdotseq {\color {Red}288.2}$

Thus we got answer. We should put the 2px width arrow (this -> ) at 288.2px position.

Position coordinate is from the left (0 px) to the right (300 px).

4. See also

Effort for reader-friendliness for general readers

Wikipedia_talk:WikiProject_Medicine/Archive_97#Infobox_update_April_2017

Introductory gene textbook website by National Library of Medicine. It includes gene location data for each gene pages.

Genetics Home Reference - gene list

5. Test with Module

at Module:Infobox gene/sandbox2

{{#invoke:Infobox gene/sandbox2|getTemplateData}}

https://en.wikipedia.org/w/index.php?title=Sonic_hedgehog&diff=prev&oldid=795122559

Category:Pages with script errors - Article namespace

6. On the width of the marker

Human
Basepairs	Approx. width in Chr.1 (Longest chromosome)	Approx. width in Chr.21 (Shortest chromosome)

248.9Mb (Chr.1 length)	300px	1599px
46.7Mb (Chr.21 length)	56px	300px
5Mb	6px	32px
2.3Mb (Longest human gene length)	2.8px	15px
1.6Mb	2px	10.6px
0.93Mb	1.1px	6px
0.31Mb	0.38px	2px
10kb (Typical human gene length)	0.012px	0.064px

Mouse
Basepairs	Approx. width in Chr.1 (Longest chromosome)	Approx. width in Chr.19 (Shortest chromosome)

195.4Mb (Chr.1 length)	300px	954.4px
61.4Mb (Chr.19 length)	94.3px	300px
3.9Mb	6px	19px
2.3Mb (Longest? mouse gene length)	3.5px	11.2px
1.3Mb	2px	6.3px
1.2Mb	1.8px	6px
0.4Mb	0.6px	2px
10kb (Typical mouse gene length)	0.015px	0.049px

Gene length is varies from one by one. So marker width also has to change page by page. Here I list up some data which are needed to think about marker width.

As result of some experiments, marker width must be at least 2px, because 1px marker is difficult to detect.

And largest marker width would be 15px (red area in the table at the right).

Length of the longest human genes are... http://www.cshlp.org/ghg5_all/section/gene.shtml

Dystrophin - 2.3 Mb at Chr.X
CNTNAP2 - 2.3 Mb at Chr.7
PTPRD - 2.3 Mb at Chr.9
RUNX1 - 1.2 Mb at Chr.21
LARGE - 760 kb at Chr.22

1px width marker

2px width marker

3px width marker

4px width marker

5px width marker

6px width marker

7px width marker

8px width marker

Red rectangle series
1x18 2x18 3x18 4x18 5x18 6x18 7x18 8x18 9x18 10x18 11x18 12x18 13x18 14x18 15x18 16x18

7. Ideograms

Human chromosome ideograms in svg
1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	X	Y

Currently used ideogram set is as above. If you want use different ideogram set, following 5 conditions must be met.

24 images are needed. (1-22 and XY)
All 24 images must have same image size (same height and same width).
Among all 24 images, pter (terminus of the p-arm, leftist point) and qter (terminus of the q-arm, rightest point) must be set at the same position.
Banding pattern must be drawn in basepair-proportional style. Standard ideograms defined by ISCN are drawn based on actual visual appearance of stained chromosomes under microscope, is not basepair-proportional. (see the table below)
All file names must have same format, changing only in chromosome number. For example, if you created chromosome 1 image named MyPrettyNice_Chr1_Ideogram.png, then the rest of file names should be as follows.

MyPrettyNice_Chr2_Ideogram.png

MyPrettyNice_Chr3_Ideogram.png

MyPrettyNice_Chr4_Ideogram.png

....

MyPrettyNice_Chr9_Ideogram.png

MyPrettyNice_Chr10_Ideogram.png

....

MyPrettyNice_Chr22_Ideogram.png

MyPrettyNice_ChrX_Ideogram.png

MyPrettyNice_ChrY_Ideogram.png

After these 5 conditions are met, you can switch current images into new images, by changing the part of the code where ideogram file name is defined.

2 types of ideogram (we should use bottom one)
Ideogram	Description	Image	The common	Difference	Cause
N We can not use this type	Chr.7 ideogram of ISCN standard, which is drawn based on actual visual appearance of stained chromosome under microscope.		In both images, band order is the same. You can see that band color, from the left to right, is set in the following order... white light grey white black white black .... This order is the same.	The order is the same. But widths of each bands are different. The salient parts are highlighted in the image below. The bands which are connected to each other are same band. You can see the difference of their width between the upper and the lower ideogram.	The cause of this difference is that basepair-density is not homogeneous within the chromosome. In some part basepairs are densely packed, and in other part basepairs are sparsely packed.
Y We can use this type	Chr.7 ideogram drawn in basepair-proportional style. As far as I know, all genome browsers (e.g. Ensembl, UCSC and so on) use this style of ideograms.

8. Forward and Reverse strands

https://www.biostars.org/p/210929/

https://www.biostars.org/p/3908/

http://seqanswers.com/forums/showthread.php?t=39388

In GRCh, as convention, direction from p-arm (short arm) to q-arm (long arm) is forward. The opposite direction is reverse.

From Nelson, Sarah C., et al. Trends in Genetics 28.8 (2012): 361-363.[1]

In all human reference chromosomes, as for other eukaryotes, the plus (+) strand is defined as the strand with its 5' end at the tip of the short arm (Genome Reference Consortium, personal communication, March 27, 2012).

Forward strand