Monday, February 23, 2015

Publish or Perish

Most of my biology-focused posts have been about topics related to garden projects I've done or have been thinking about. The subject of gardening does account for much of the biology I spend my free-time thinking about, especially now in the middle of a winter-bound Minnesota. It is also far easier to examine the growth and interactions of living things in the garden than it is to go on expeditions or develop lab protocols to explore the diversity of biological topics.

I've avoided talking in detail about the biology I do at work because the work is on-going and to publicize it here might detract from the process of publishing the work in more formal contexts once the projects reach usable conclusions. Sometimes a key result can be quickly replicated once the idea behind it has been developed, so talking about results too early can be asking for some competitor to beat you to publication. I've defended my thesis and dealt with graduate school bureaucracy sufficiently to be awarded my PhD. I can now refer to myself as "Darren Abbey, PhD" in professional contexts. (I can also refer to myself as "The Doctor" (Dr. Who?) in certain social contexts.)

Along the way, I've written or contributed to several research publications. The following list is the publications I've been received name credit for my contributions, from oldest to most recent.

  1. Gale CA, Leonard MD, Finley KR, Christensen L, McClellan M, Abbey D, Kurischko C, Bensen E, Tzafrir I, Kauffman S, Becker J, Berman J. (2009) SLA2 mutations cause SWE1-mediated cell cycle phenotypes in Candida albicans and Saccharomyces cerevisiae. Microbiology. 155(Pt 12):3847-59. [PMID: 19778960]
     
  2. Forche A, Abbey D, Pisithkul T, Weinzierl MA, Ringstrom T, Bruck D, Petersen K, Berman J. (2011) Stress alters rates and types of loss of heterozygosity in Candida albicans. MBio. 2(4). [PMID: 21791579]
     
  3. Abbey D, Hickman M, Gresham D, Berman J. (2012) High-Resolution SNP/CGH Microarrays Reveal the Accumulation of Loss of Heterozygosity in Commonly Used Candida albicans Strains. G3 (Bethesda). 1(7):523-30. Erratum in: G3 (Bethesda). 2(11):1473. [PMID: 22384363]
     
  4. Hickman MA, Zeng G, Forche A, Hirakawa MP, Abbey D, Harrison BD, Wang YM, Su CH, Bennett RJ, Wang Y, Berman J. (2013) The 'obligate diploid' Candida albicans forms mating-competent haploids. Nature. 494(7435):55-9. [PMID: 23364695]
     
  5. Abbey DA, Funt J, Lurie-Weinberger MN, Thompson DA, Regev A, Myers CL, Berman J. (2014) YMAP: a pipeline for visualization of copy number variation and loss of heterozygosity in eukaryotic pathogens. Genome Med. 6(11):100. [PMID: 25505934]
     
  6. Ford CB, Funt JM, Abbey D, Issi L, Guiducci C, Martinez DA, Delorey T, Li BY, White TC, Cuomo C, Rao RP, Berman J, Thompson DA, Regev A. (2015) The evolution of drug resistance in clinical isolates of Candida albicans. Elife. 4. [PMID: 25646566]
     
I've also got a couple others in the pipeline. For one I'm looking for a target journal, for the other I'm still deciding exactly how to present the results. I'll let you know more details once they're further along the way to publication.

The projects a grad student works on depends on a mix of the lab they end up in and their personal style of problem solving. I ended up in a Candida albicans lab and brought to it a heavy computational approach. The mix between the two is the realm of computational biology, the topic I find myself most connected to.

The academic life can readily be described as, "Publish or Perish". With six name papers from my time in grad school, I've done alright. Now I just have to figure out the next step.

Thursday, February 19, 2015

The Color of Cotton

1. www.perunaturtex.com/scientif.htm
Ever since I first heard about it, I've been interested in naturally colored cotton (also known as "color grown cotton"). The cotton plants most people are familiar with produce a pure white fiber, which can then be dyed to match any desired color. Naturally colored cotton, on the other hand, is grown with color straight from the plant.

There is archaeological evidence for the existence of cotton in various shades of yellow, brown, green, and red. The pre-Columbian Peruvian textile at left is supposedly made from colored cotton without the processing of additional dyes. There are also reports of a naturally colored blue cotton, but the internet has provided minimal evidence for this.

The majority of commercially grown cotton belongs to the species Gossypium hirsutum. Varieties of this cotton species can be found in light brown and green, but the other colors are generally nowhere to be found.

2. Inheritance of different fiber colors in cotton.
During a recent web search, I came across an image of a very dark brown naturally colored cotton. The image also shows a nice orange cotton, in addition to the more typically seen tan and light brown colors. Seed for these varieties aren't generally available, however, but they can be accessed from different seed collections (such as GRIN) if you can show you have some worthwhile research, education, or other public good rationale for having them.

These more interesting colors are seen in relatively wild varieties of a second cotton species, G. barbadense. G. hirsutem and G. barbadense don't naturally cross in the wild due to different timing of pollen maturation and other mechanisms. These are all easily circumvented by directed hybridization efforts. The resulting F1 hybrids grow well and are fertile, though fertility issues do arise in some F2 generation plants. These issues wouldn't interfere greatly with the intentional recombination of pigment alleles from both species, which I think has interesting potential to create new interesting colored varieties.

After I've done some further experiments with growing the cotton lines I already have in Minnesota, I might request some of the interesting colored forms from the collections.

3. Cotton and I.
...wait, growing cotton in Minnesota (red star)? ...a thousand miles north of where cotton is grown in the USA (green regions)?

A few years ago, on a whim, I planted some cotton seeds I had come across in south Texas. Cotton is typically described as needing a long and hot growing season to mature. I don't know how the production of my plants compared to similar plants in the South, but their production was dramatically higher than I expected up here in the North. After the first hard freeze of winter, I broke the plants off at their base and hung them to dry in the garage. A few weeks later, the cotton fibers had completed drying and were easy to collect. This process might not be something that can be scaled up to a proper crop, but it might. Further research is needed.

References:
  1. www.perunaturtex.com/scientif.htm
  2. Dark brown cotton: www.scielo.br/scielo.php?pid=S1984-70332014000400008&script=sci_arttext
  3. GRIN: www.ars-grin.gov/npgs/orders.html
  4. Crossing: www.ogtr.gov.au/internet/ogtr/publishing.nsf/content/cotton-3/$FILE/biologycotton08.pdf

Thursday, February 12, 2015

The Case of the Marvelous Meiosis

1. Meiosis in male (A) and female (B).
The typical story of meiosis is described in detail in many a biology textbook. It all starts with a diploid somatic cell. The chromosomes duplicate and homologs align along the cell's mid-line. The aligned chromosome doublets crossover and then the doublets are divided into two new cells (Meiosis I). The chromosome doublets align along the mid-lines and the new cells divide again (Meiosis II) to produce the final gametes. This model well matches the process of spermatogenesis (Fig 1A), which was used as a model to study and understand the process. Oogenesis follows basically the same process, except that each division is asymmetric. The result is one large gamete (Fig 1B) and two or three non-viable cell fragments with the extra chromosomes.

2. Meiosis in male (A) and female (B) Caninae group roses.
Roses in the group Caninae (Dog Roses) do a peculiar version of the process. The basic example is tetraploid, but produces haploid male (Fig 2A) and triploid female (Fig 2B) gametes. During meiosis I, two homologs for each chromosome align to form a bivalent and the rest remain as monovalents. The monovalents are lost during spermatogenesis, but retained during oogenesis. This system is referred to as "permanent odd polyploidy".  It even works with extra uneven chromosome copies. The extras form monovalents and are discarded or retained just like the red and blue chromosomes in Figure 2. This system allows fertility to be maintained even with odd chromosome counts that would normally make a plant sterile.

I came across this peculiar variant of meiosis while researching what complications might develop in a cross between Rosa pomifera (4n, group Caninae) and R. rugosa (2n). Both parent species have relatively large fruit. My initial plan was to hybridize the two species, generating at least one F1 plant, then allow them to self and screen many F2 progeny plants for increased fruit size.

That the two roses are not very closely related and both species have numerous smaller-fruited relatives suggests their large fruit was evolved separately. This is useful for a breeding project as it means that different mutations accumulated as each plant developed large fruit. Those separate mutations can possibly be recombined via hybridization to generate a progeny plant with even larger fruit.

I expect the project of domesticating roses as [more of] a fruit will take decades. It should be an amusing hobby and I expect to have enough time left to see some nice results.

References:
  1. Meiosis: en.wikipedia.org/wiki/Meiosis
  2. Rose groups: en.wikipedia.org/wiki/List_of_Rosa_species
  3. Dog rose meiosis:
    1. www.nature.com/hdy/journal/v101/n4/abs/hdy200863a.html
    2. www.ncbi.nlm.nih.gov/pubmed/24685720

Wednesday, February 4, 2015

Mathematical Recreations : Homeomorphically Irreducible Trees of size n=10

A homeomorphically irreducible tree is an acyclic graph where there are more than two branches from each internal vertex. The size n=10 means there are ten vertices, internal or edge, in the tree.

A. Homeomorphic irreducible trees of n=10.
In the movie "Good Will Hunting", the puzzle of calculating all of such trees with a total of ten vertices was presented to a class as a difficult challenge that took a couple years to calculate.  The puzzle came to my attention after watching a Numberphile video on youtube, where the guest mathematician rightly pointed out that it really isn't that difficult of a puzzle as he drew out all ten possibilities.

There are lots of blog posts pointing out how simple the problem actually is, but neither the Numberphile mathematician or any of the various bloggers point out how the ten possibilities are connected by a set of very simple transformations.

B. Transformation rules.
The first transformation (fig B1) takes three leaves (edge vertices) from an internal vertex and generates a new internal vertex with two leaves.

The second transformation (fig B2) moves a leaf from one internal vertex to another.

You can't apply a transformation such that it will generate a tree with only two branches from an internal vertex, as the resulting tree will no longer be homeomorphically irreducible.

C. Transformation relationships.
With these transformations, it is a simple task to generate all the possible homeomorphically irreducible trees for higher numbers of vertices. (At least for every case that I've examined.) You start with the simplest case of the pinwheel and apply as many of the transformations as you can at each step until you run out of steps. Proving that this process will generate all possibilities for any number of vertices is a more difficult challenge.

I suspect there might be a set of such trees where individuals are isolated from others, where another transformation would be required.

D. Transformation graph.
The map of trees connected by transformations generates a new graph (such as fig D). My limited scribbling of the trees for a set of 11 and 12 nodes shows that there is more than one valid transformation graph for some sets of trees. I'll have to explore the properties of these transformation graphs further at some point.