sleep_analysis

Streamilt App To-DO

• Main Settings Page:
  • Locomotor_activity_by_day.csv
  • Nighttime_daytime_activity_ratio.csv
• Sleep Analysis
  • Individual_day_night_sleep.csv numbers are slightly off

Shiny App

• https://karolcichewicz.shinyapps.io/shinyr-dam/
• git - https://github.com/KarolCichewicz/ShinyR-DAM/

Streamlit App

• use rnaseq virtual environment

Locomotor activity

1. Put the settings exactly like in Main_settings.png
2. Run and make sure numbers match Locomotor_activity_mean.png
3. Use Monitor9.xlsx to see how we got it
4. Should be 29 alive flies, 3 dead flies, and a mean of 10190.11

Sleep_analysis_settings

1. Make sure settings match Sleep_analysis_settings.png

Psuedo-code for initial upload and statistics

• Read in the monitor file and give the following column names when read in file or create dataframe:
• Index,Date,Time,LD-DD,Status,Extras,Monitor_Number,Tube_Number,Data_Type,Unused,Light_Sensor,data_1,data_2,data_3,data_4,data_5,data_6,data_7,data_8,data_9,data_10,data_11,data_12,data_13,data_14,data_15,data_16,data_17,data_18,data_19,data_20,data_21,data_22,data_23,data_24,data_25,data_26,data_27,data_28,data_29,data_30,data_31,data_32
• Set variables - LD_start = 06:00, counts_per_day_alive = 100, LD_DD_anlysis = 'LD', start_date = 2/24/2024, end_date = 2/27/2024
• Create column called LD_DD with LD_start + 18:00 = LD else DD
• Filter on date between start_date & end_date
• Filter on LD_DD based on LD_DD_analysis
• Create column called date_LD_DD based on concat of date + LD_DD
• Group by date_LD_DD and sum each column of data_1 - data_32
• Loop through data_1 .. data_32 and determine the alive flies & dead flies - any fly with less than counts_per_day_alive is removed from the analysis - remove from the dataframe and keep number_of_dead_flies
• Calculate number of alive flies, number of dead flies, Mean, SD, SEM - stats only on alive flies
• Make sure the output is the same as the shiny app

Excel logic

• read in file - 7352 rows
• filter by date - 24-27 - now have 5760 rows
• filter by time and then call everything before 6:00 DD, everything between 6:00 - 18:00 LD, 18-24 is DD
• filter where LD_DD = LD --> 2880 rows
• make sure the Monitor_Number column is correct - might be adding support for multiple monitor files together

PSUEDO-CODE FOR SLEEP ANALYSIS

Question for Alyssa - do you want to be able to use previous monitor runs - should the software keep track of all of your analyses

Next Steps

1/10/2025

• set settings page to wide layout and have 3 columns with the inputs

• need to decide on graphs on main page

• moved Settings_and_daily_locomotor_activity.py to Home.py (outside of pages)

• moved the Sleep_Analysis.py into the pages as 0_Sleep_Analysis.py

• Sleep_Analysis.py page ** Daily_sleep_profiles.csv *** go to previous notebook and see if just take average at each min instead of average over the entire period

  Dec_time Condition date Dec_ZT_time mean_of_sleep_counts sem 1 0 wt 2/24/2024 1080 0.724137931 0.084465164 2 1 wt 2/24/2024 1081 0.689655172 0.08742975

** Average_sleep_profiles_in_LD.csv *** sorting the previous dataframe by Dec_time (includes multiple days at same Dec_time) and then binning based on the user input (bin sleep profile with default = 30) averaging mean of sleep counts *** 0-29, 30-59, 60-89, ...

** Individual_day_night_sleep.csv *** need to cap all values to 1.2 if they are greater

** Individual_sleep_activity_bout_data.csv

PSEUDO-CODE