Open lcolladotor opened 4 years ago
ggplot2
y cowplot
en vez de plot()
https://github.com/lcolladotor/osca_playground_leo/commit/5868d024ca741eb516fef513bf5ded13a171288d. @AnaBVA mostró otra versión usando también ggplot
(tal vez vivirá en https://github.com/AnaBVA/scRNAseq). Ver https://github.com/lcolladotor/osca_playground_leo/blob/master/ERCC_example_ggplot2_version.pdf. table()
, which()
e intersect()
usando los operadores lógicos |
(or), &
(and) y !
(not). Expandimos las respuestas en https://github.com/lcolladotor/osca_playground_leo/commit/38f8dd55cc861602b06aa30572a632c7e31cf5e7(edit) En algún momento del día también checamos http://bioconductor.org/packages/release/bioc/html/SingleCellExperiment.html y su relación con http://bioconductor.org/packages/release/bioc/html/SummarizedExperiment.html.
Topic: scRNA-seq with Leonardo - Zoom Meeting Start Time : Mar 25, 2020 10:59 AM
Meeting Recording: https://jh.zoom.us/rec/share/utR_NZ392EhLeKf16Gz4VPIdJZ27aaa8gCNM-_ZYn0xVpNsed1hCYcblhwPb2pYi
emptyDrops()
siguiendo parte del artículo https://www.ncbi.nlm.nih.gov/pubmed/30902100.+52 554 161 4288
. Esto le ayudó a unos para que siguieran nuestra conversación de audio vía el teléfono y ocasionalmente pudieran ver la pantalla cuando su conexión de internet se los permitió. colorout
para que su terminal les muestre el output de R en colores. Lo pueden hacer en todas sus sesiones de R de forma automática usando código como este en su archivo de ~/.Rprofile
:## Change colors
# Source https://github.com/jalvesaq/colorout
if(Sys.getenv('TERM') %in% c("term", "xterm-256color", "cygwin", "screen") & interactive()) {
installed <- require("colorout")
if(!installed) {
cat('devtools::install_github("jalvesaq/colorout")\n')
}
rm(installed)
}
b
dentro de la ecuación de a * b = c
y resolvimos más ejercicios en https://github.com/lcolladotor/osca_playground_leo/commit/d1affa8e171a15e373578e4c645e68dd8944984d. Esta vez, por el tiempo, no trabajamos en pares. Si no que fuimos rotando quién mostraba su pantalla (excepto para quienes tenían problemas con su conexión a internet).Topic: scRNA-seq with Leonardo - Zoom Meeting Start Time : Mar 26, 2020 11:01 AM
Meeting Recording: https://jh.zoom.us/rec/share/vMZ0KbLU_D5OQJHR2Hjmc7MbLraiT6a8hnRN-fVcyxwERmYmkf0aYOVdgUmzNvxG
Esta vez también grabé localmente cuando hicimos la actividad de los "breakout rooms" (zoom no graba en la nube cuando no estamos en el grupo principal basado en https://kb.iu.edu/d/arwb). Aquí está esa grabación: https://www.dropbox.com/sh/go4soylfq1nbzb4/AAB3yrzP4jjvC2IuRADdAR7da?dl=0.
Reanudamos corriendo el código de ayer de https://github.com/lcolladotor/osca_playground_leo/blob/master/R/04-normalization.R#L1-L72 mientras repasamos las diapositivas de https://lcolladotor.github.io/osca_LIIGH_UNAM_2020/04-normalization.html que ya habíamos visto. @ Lucia (no me sé su usuario de GitHub) y @avleon sugirieron usar screen
para no tener que volver a correr el código en caso de perder la conexión al cluster aunque tendrían que volver a hacer el mount y todo eso.
Hablamos sobre tener cuidado al ver gráficas de puntos y buscar si es que enmascaran dos o más lineas como en el siguiente ejemplo.
Vimos detalles de logNormFactors()
y ecuaciones relacionadas tipo: counts / lib.sz.factor
, log(counts / lib.sz.factor + x)
, x = 1
, log(0) = -Inf
, log(1) = 0
.
En relación al ejercicio de dos genes
hablamos sobre la ecuación para definir una prueba estadística de valor t
https://en.wikipedia.org/wiki/T-statistic
y sobre la desviación estándar https://en.wikipedia.org/wiki/Standard_deviation
Les recomendé los cursos de https://leanpub.com/universities/set/jhu/cloud-based-data-science que pueden tomar de forma gratuita por si quieren aprender más de R. Los hizo mi jefe del doctorado con su equipo.
Hicimos los ejercicios de https://lcolladotor.github.io/osca_LIIGH_UNAM_2020/05-feature-selection.html#9 en pares de equipos de 2 o 3 personas. Luego cada alumno explicó una o más líneas de código de https://github.com/lcolladotor/osca_LIIGH_UNAM_2020/blob/master/05-feature-selection.R#L57-L105 como su examen final del curso. Les aplaudimos a todos los alumnos =) muchas veces con el 👏🏽👏🏽👏🏽en Zoom.
Revisamos otra vez el ejemplo inicial de la introducción en https://lcolladotor.github.io/osca_LIIGH_UNAM_2020/01-introduction.html#31 y https://lcolladotor.github.io/osca_LIIGH_UNAM_2020/01-introduction.html#32, y aunque algunas preguntas eran sobre temas que no nos dio tiempo de ver, creo que los alumnos entendieron bien los pasos que si vimos. Así que nos quedamos a la mitad de la clase 5 de las 11 originalmente planeadas en https://lcolladotor.github.io/osca_LIIGH_UNAM_2020/ sobre scRNA-seq.
Después del descanso, les platiqué de nuestro pre-print reciente de Spatial Transcriptomics (el primero de 10x Genomics Visium)
shiny
para este proyecto: http://spatial.libd.org/spatialLIBD/spatialLIBD
(que expande el objeto de SingleCellExperiment
): http://research.libd.org/spatialLIBD/regutools
y a spatialLIBD
: https://github.com/Bioconductor/Contributions/issues?page=1&q=is%3Aissue+is%3AopenspatialLIBD
https://github.com/Bioconductor/Contributions/issues/1389Les platiqué sobre mi promoción reciente a Research Scientist en LIBD http://lcolladotor.github.io/2020/03/16/research-scientist-an-academic-career-launch-pad/ y mis intereses de investigación que viene escrito en http://lcolladotor.github.io/authors/admin/ y en mi CV http://lcolladotor.github.io/cv/. También les aclaré que en LIBD tenemos un programa de estancias de verano pagadas (aunque no podemos dar visas para esas estancias) y que próximamente empezaré a armar mi propio grupo (ahí si puedo contratar a gente de México o de donde sea).
Les platiqué sobre mi carrera haciendo referencias a http://lcolladotor.github.io/2019/04/10/the-evolution-of-my-academic-career-as-seen-through-posters-and-talks-thanks-to-hugo-academic-4-1/#.Xn6qGtNKhZ0 y http://lcolladotor.github.io/2018/11/06/a-knot-of-threads-from-cshl-to-lcg-unam-to-aldo-barrientos-to-diversity-scholarship-opportunities/#.Xn6qdNNKhZ0 además de mostrar diapositivas de https://speakerdeck.com/lcolladotor/cdsbmexico.
Les recordé que siempre nos pueden pedir ayuda y participar en la CDSB ya que varios ahí estamos interesados en ayudar a otros para que tengan las misma oportunidades que tuvimos (o aún mejores). Por ejemplo, si quieren mandar una solicitud para un congreso en EEUU y no lo han hecho antes, les podemos dar sugerencias y demás via un archivo de Google Doc.
Les pedí que me avisen si tienen una cuenta de Twitter académica para que pueda seguirlos y ver como van creciendo y demás.
Me preguntaron por Slack sobre el iGEM y les dije que yo no participé en el iGEM durante la LCG pero que Mariana Gómez si fue una de las líderes del proyecto de mi generación en la LCG-UNAM https://twitter.com/mgschiavon.
Topic: scRNA-seq with Leonardo - Zoom Meeting Start Time : Mar 27, 2020 10:58 AM
Meeting Recording: https://jh.zoom.us/rec/share/58hvFJLC8T9LEp3ns0aCU658AtXvT6a823MY86ZexRwlb3BuVIGy8h2NZ3dsrepa
y https://www.dropbox.com/sh/evdhe518skot8fb/AADtwX543Et9SeKu3jLK7QzFa?dl=0
Día 1: mañana
sshfs
los archivos del cluster en su laptop (gracias a @Lucía Ramírez). En ambos casos tienen que usar la terminal de RStudio para correr R en el cluster. Alternativamente, pueden copiar y pegar código a una terminal corriendo R del cluster.Día 1: tarde
pdf('nombre_de_archivo.pdf'); plot(); dev.off()
o habilitar X11 (como les explicó @Lucía Ramírez o editando su archivo de~/.ssh/config
en su laptop) antes de usarssh
para conectarse al cluster.SingleCellExperiment::SingleCellExperiment()
y respondimos unos ejercicios https://github.com/lcolladotor/osca_playground_leo/commit/31005f68320a27c3a327cf063aa3c260b2305960. Nos quedamos en la diapositiva 51 de Pete https://docs.google.com/presentation/d/1X9qP3wNlnn3BMUQhuZwAo4vCV76c33X_M-UnHxkPZpE/edit#slide=id.g7e63e0fe24_0_151 y en la diapo 5 de las mías https://lcolladotor.github.io/osca_LIIGH_UNAM_2020/02-data-infrastructure-and-import.html#11Grabación
Topic: scRNA-seq with Leonardo - Zoom Meeting Start Time : Mar 24, 2020 10:45 AM
Meeting Recording: https://jh.zoom.us/rec/share/2NNsaJXv30JJbK_V-nnWR4s6M734X6a813UW-aUKmhsmvXdv8AoC15_QeH4vwSK6