Riba:
- Ikon ƙirƙira tsarin aiki na matakai bakwai daga tambaya zuwa tabbataccen sakamako ta hanyar sanya ƙofar tabbatarwa a kowane mataki.
- Ikon tabbatar da lambar Python da aka rubuta ta hanyar basirar wucin gadi tare da sanannun bayanan gwaji da kuma bambanta bambanci tsakanin 'lambar aiki' da 'madaidaicin lambar'
- Sanya binciken ya zama mai sake fasalin (siffa, matsakaici, iri, daftarin aiki) kuma bayar da rahoto tare da rigar tabbaci, bayyana gaskiya da amincewar ƙwararru
Mun koyi guntuwar a cikin raka'o'i tara da suka gabata: bincike na jeri, omics, ƙirar furotin, ƙirar gwaji, bayanan dakin gwaje-gwaje, sarrafa halittu, adabi, da ɗa'a. A cikin wannan naúrar ta ƙarshe, za mu haɗa guda ɗaya tare da gina aikin aiki na ƙarshe zuwa ƙarshen - sarkar daga tambayar bincike zuwa ƙarshen ingantaccen, inda AI ke taimakawa a kowane mataki amma ɗan adam yana yin kowane yanke shawara mai mahimmanci da tabbaci. Har ila yau, za mu rufe Python, wanda shine kashin bayan wannan sarkar, da kuma horo na reproducibility - ikon maimaita bincike ta wani, tare da bayanai iri ɗaya, don ba da sakamako iri ɗaya.
Manufar ba shine a ba da kayan aikin ɗaiɗaikun ɗaiɗaiku ba, amma tunani mai alhakin tafiyar aiki: zaku san inda za'a saka AI, inda za'a sanya ƙofar tabbatarwa, da yadda ake iya gano tsarin gaba ɗaya.
Me yasa Python?
Harshen yare na bioinformatics da ilimin lissafin lissafi shine Python (da R). Domin zaku iya sarrafa kansa da yin kusan kowane mataki maimaituwa tare da buɗe ɗakunan karatu na tushen (Biopython don nazarin jerin abubuwa, pandas don tebur ɗin bayanai, scikit-koyi don koyon injin, matplotlib don gani). AI yana da ƙarfi sosai wajen rubuta lambar Python; Amma karɓar lambar da ta samar ba tare da aiki ba kuma tabbatar da shi yana da haɗari - lambar na iya mayar da sakamakon da ba daidai ba cikin shiru (kuskuren naúrar, ginshiƙi mara kyau, tacewa da aka rasa).
Hankali: Ko da lambar da AI ta rubuta tana aiki, mai yiwuwa ba daidai bane. "Ya yi aiki ba tare da kurakurai ba" da "ya ba da sakamako daidai" abubuwa biyu ne daban-daban. Gwada kowace lamba tare da ƙarami, sanannun bayanan gwaji: shine shigarwa-fitarwa kamar yadda kuke tsammani? Wannan ita ce hanya ɗaya tilo don kama kurakuran shiru.
Anatomy na aikin aiki
Alhaki mai alhakin AI mai kunna aikin injiniyan halittu yana bin wannan tsarin:
- Tambaya da hasashe. Tambaya bayyananne, mai iya gwadawa. (AI na iya ƙarfafawa; kun bayyana.)
- Tarin bayanai da sarrafa keɓantawa. Ina bayanan daga, na sirri ko amintaccen kafofin watsa labarai? (raka'a 9)
- Pre-processing da ingancin iko. Tsaftacewa, daidaitawa, sarrafa tsari. (Raka'a 2-3.)
- Bincike. Ƙididdiga, ƙirar ƙididdiga, ƙididdiga. (Ƙofar Tabbatarwa a kowane mataki.)
- Sharhi. Buga tunanin nazarin halittu, bambance-bambancen dalili.
- Tabbatar da dakin gwaje-gwaje jika. Ana gwada hasashe mai mahimmanci da gwaji. (Kashi na 1, 4, 5.)
- Bayar da rahoto da bayyana gaskiya. Lambar da za a sake bugawa, bayanin AI, amincewar ƙwararru. (Kashi na 8-9.)
Kowane mataki yana da wurin bincike - wurin da aka tabbatar da fitarwa kafin a ci gaba zuwa mataki na gaba. AI yana haɓaka mataki ɗaya, ba za ku iya ci gaba zuwa mataki na gaba ba tare da wucewa kofa ba.
Tukwici: Ajiye aikin ku azaman rubutun da littafin rubutu; Bari a rubuta shigarwar, fitarwa da yanke shawara na kowane mataki. Samun damar amsa tambayar "ta yaya na sami wannan sakamakon" a cikin 'yan mintoci bayan watanni yana da darajar zinari ga duka kimiyya da dubawa.
Reproducibility: inshora na sakamakon
Sakamakon abin dogara ne kawai idan wani zai iya maimaita shi. ginshiƙai huɗu na sake fasalin su ne: (1) code yana cikin sarrafa sigar (tare da git), (2) an gyara muhalli (an yi rijista nau'ikan ɗakin karatu, misali buƙatun.txt ko muhallin conda), (3) an yi rajistar bayanai da iri bazuwar, (4) an rubuta kowane mataki. AI yana taimakawa gina wannan ababen more rayuwa, amma ya rage naku don aiwatar da horo.
uku mini lokuta
Case 1 - An kama kuskuren lambar shiru. Wata ƙungiya ta yi amfani da rubutun daidaitawa wanda AI ya rubuta; Lambar tana aiki ba tare da kurakurai ba. Lokacin da suka gwada shi tare da sanannun bayanan gwaji, sun gano abin da ake fitarwa ya canza ta hanyar 1,000 - rubutun yana yin juzu'in juzu'in naúrar da ba daidai ba. Ƙananan bayanan gwajin sun sami kuskuren da zai rushe duk binciken.
Hali na 2 - An adana sakewa. Bayan an watsa wasan, alkalin wasa ya bukaci a sake buga sakamakon. Tawagar ta sake fitar da tantancewar a zahiri cikin mintuna 20 saboda suna da sigar lambar a Git da mahalli mai maƙalli. Ƙungiya mai hamayya da ba ta yi rikodin yanayin ba ta kasa cimma wannan buƙatun tsawon makonni.
Case 3 - Tabbatarwa daga ƙarshe zuwa ƙarshe. A cikin wani aikin enzyme, aikin aikin ya yi aiki kamar haka: AI ya ba da shawara daga wallafe-wallafen (wanda aka tabbatar), samfurin sunadaran sunadaran maye gurbi (an gwada su ta hanyar gwaji), DoE ya rage yawan gwaje-gwajen, daya daga cikin uku na maye gurbi ya karu da aiki ta hanyar 1.9-ninki. AI ya haɓaka a kowane mataki, an tabbatar da ɗan adam a kowace kofa; Sakamakon ya kasance duka cikin sauri kuma mai karewa.
Samfura huɗu masu kwafi
1) Ƙaddamar da kwarangwal na aiki:
Matsayinku: Mashawarcin aikin nazarin halittu na lissafi. Ƙirƙirar aikin aiki na ƙarshe zuwa ƙarshe don amsa tambaya mai zuwa: [tambaya]. Ga kowane mataki, (1) abin da za a yi, (2) inda AI ke taimakawa, (3) abin da tsarin tabbatarwa ya kamata ya zama, (4) wane kayan aiki don amfani. Haɗe rigar ingantaccen lab da matakin bayyana gaskiya.
2) Python code + buƙatar gwaji:
Matsayinku: mai haɓaka bioinformatics. Rubuta aikin Python wanda ke yin aiki mai zuwa: [aikin]. Laburare: [pandas/Biopython]. Hakanan ƙara ingantaccen misali tare da sanannun bayanan gwaji: menene shigarwar, menene fitarwar da ake sa ran. Zan gudanar da lambar kuma in duba shi tare da bayanan gwaji. Aiki maras wanzuwa/daidaita siga.
3) Duban sake haifuwa:
Ina so in mayar da bincike na. Ba ni jerin abubuwan dubawa: sarrafa sigar, mahalli (bukatun buƙatun/conda), iri bazuwar, rajistar tushen bayanai, takaddun mataki. Ba da hanya mai amfani don kowane abu.
4) Tabbatar da sakamakon:
Ina so in yi gwajin tabbatarwa na ƙarshe kafin sanya sakamakon bincike a cikin rahoto. Ba ni lissafin waɗannan tambayoyin: shin sakamakon yana da kyau a cikin ilimin halitta, shin ƙididdiga daidai ne (gwaji da yawa, samfuri), an tabbatar da shi ta hanyoyi masu zaman kansu, shin ya dogara da tushe, ana buƙatar gwajin rigar, an yi bayanin AI?
Rauni mai ƙarfi / Ƙarfi mai ƙarfi
Rawanin faɗakarwa:
Rubuta mani lambar Python mai nazarin wannan bayanan.
Muhimmin manufa, babu gwaji, babu tabbaci; kuskuren shiru mai saurin faruwa.
Ƙarfi mai ƙarfi:
Matsayinku: mai haɓaka bioinformatics. Don CSV (ginshiƙai: gene, control_mean, treatment_ma'ana, pvalue) tare da pandas: (1) ƙara ginshiƙan canjin log2, (2) ƙididdige ƙimar BH da aka gyara, (3) tace padj<0.05 da | log2FC|>1. Hakanan nuna fitowar da ake tsammanin tare da layuka 5 na bayanan samfurin don in tabbatar. Kar a yi amfani da sunan ginshiƙi mara kyau ko aikin da ba ya wanzu.
Bambanci: bayyanannen manufa, bayyananniyar ƙididdiga, tabbatarwa tare da bayanan gwaji da hana ƙirƙira.
Ƙofar tabbatar da kwararar aiki daga ƙarshe zuwa ƙarshe
mataki
AI gudunmawa
kofar tabbatarwa
hasashe
wahayi, adabi
Shin ana iya gwadawa ko kuma aka yi masa walda?
Bayanai / sirri
jerin abubuwan dubawa
De-gano, ingantaccen muhalli
preprocessing
rubutun
Batch/QC iko
Bincike
code, model
Bayanan gwaji, abin hawa mai zaman kansa
Sharhi
daftarin aiki
Hankalin halittu, alaƙa-dalili
rigar tabbaci
Shirin gwaji
Sakamakon gwaji na gaske
Rahoton
daftarin aiki
Maimaituwa, nuna gaskiya, amincewar ƙwararru
Kuskuren gama gari
- Tunanin cewa lambar aiki daidai ne. "An yi aiki ba tare da kurakurai ba" ≠ "sakamakon daidai"; yakamata a gwada tare da bayanan gwaji.
- Ketare ƙofofin tabbatarwa. Wucewa kofa don gudun gudu yana sanya duk matakan da suka biyo baya cikin haɗari.
- Yin watsi da haifuwa. Ba tare da rajistar muhalli/version ba sakamakon ba zai iya sakewa ba.
- Jinkirta / tsallake tabbacin rigar. Ba za a iya yanke shawara ba har sai an tabbatar da hasashen kwamfuta ta hanyar gwaji.
- Mantawa da bayyana gaskiya da amincewar masana. Sa hannu na ƙarshe da sanarwar AI wani ɓangare ne na tafiyar aiki.
a takaice
Mahimmancin bioengineering yana amfani da AI ba azaman kayan aikin mutum ɗaya ba, amma a matsayin wani ɓangare na aikin aiki na ƙarshe-zuwa-ƙarshe wanda aka liƙa tare da ƙofofin tabbatarwa. Python da reproducibility sune kashin bayan wannan kwarara; AI yana rubuta lamba amma kuna tabbatar da shi tare da bayanan gwaji, saboda lambar aiki ba lambar daidai ba ce. AI yana haɓaka a kowane mataki, ɗan adam yana tabbatarwa a kowace kofa; Ana gwada ƙididdiga masu mahimmanci a cikin dakin gwaje-gwaje mai jika, kuma ana ba da rahoton sakamakon a bayyane kuma tare da amincewar ƙwararru. Ma'anar wannan ƙirar tana cikin jumla ɗaya: Ɗauki saurin AI, kiyaye yanke shawara da alhaki a cikin ɗan adam.
Aikin aikace-aikace
Ƙirƙirar farawa-zuwa-ƙare aikin aiki don tambayar bincike na kanku. Shin AI ta fito da tsari na matakai bakwai tare da samfurin "kwarangwal ɗin aiki" kuma ƙara ƙofar tabbatarwa ga kowane mataki. Sannan buga rubutun tare da samfurin "Python code + test request" don ɗayan matakan bincike, gudanar da shi tare da ƙananan bayanan gwaji kuma tabbatar da cewa fitarwa daidai ne. A ƙarshe, shirya jerin "tabbacin tabbatar da sakamako" kuma a shirya wannan aikin don yin rahoto. Yi rikodin tsarin gaba ɗaya a cikin tsarin da za a iya sakewa (lambar + kafofin watsa labarai + daftarin aiki).
jerin abubuwan dubawa
- [ ] Na tsara tsarin aiki tare da matakai bakwai da ƙofar tabbatarwa a kowane mataki.
- [ ] Na tabbatar da lambar da AI ta rubuta tare da sanannun bayanan gwaji.
- [ ] Na sanya binciken ya sake sakewa (siffa, muhalli, iri, takarda).
- [ ] Na dangana mahimmin hasashe ga ingantaccen laburaren jika.
- [ ] Na gina bayanan sirrin bayanai da sarrafa lafiyar halittu a cikin aikin aiki.
- [ ] Na kammala rahoton tare da fayyace bayanan AI da amincewar ƙwararru.
Gwajin Module
1. Wanne daga cikin waɗannan shine mafi daidaiton matsayi don basirar wucin gadi a cikin injiniyoyin halittu?
- A) AI shine kayan aikin nunawa da tsarawa; Alhakin yanke shawara da ke ƙayyade ma'anar halitta da aminci yana kan mutane ✔
- B) Hankali na wucin gadi na iya sanya ƙwayoyin ɗan takara kai tsaye cikin samarwa ba tare da amincewar ɗan adam ba
- C) Tunda hankali na wucin gadi koyaushe yana da manufa fiye da mutane, ya kamata a bar shi tafsirin halittu.
- D) Hankali na wucin gadi yana da amfani kawai don taƙaita rubutu, ba shi da alaƙa da bayanan dakin gwaje-gwaje
Bayani: Hankali na wucin gadi; Mataimaki ne wanda ke bincika bayanai masu ƙarfi da hayaniya, sa alama kuma yana samar da zane-zane. Kwararre ne wanda ya yi ma'anar nazarin halittu, aminci da yanke shawara na ƙarshe; Buga da ba a tantance ba zai iya sanya majiyyaci, rukunin samarwa, ko bugawa cikin haɗari. A cikin wurare masu mahimmancin aminci, fitowar AI ba ta zama madadin amincewar ƙwararru ba.
2. Menene maƙasudin matakin 'hanyar tushen' lokacin da aka tabbatar da fitarwar AI?
- A) Kawar da ƙayyadaddun ƙirƙira (hallucinatory) ta hanyar haɗa kowane da'awar zuwa takamaiman bayanai ko tushen asali ✔
- B) Samar da fitar da ruwa mai ruwa da kuma iya karantawa
- C) Tsallake tabbatarwa don gama bincike cikin sauri
- D) Yarda da nassoshi da aka bayar da hankali na wucin gadi kamar yadda suke
Bayani: AI yana da kyau amma lokaci-lokaci yana haifar da hallucinations; na iya gabatar da kwayar halitta, hanya, ko tunani a matsayin gaske. Haɗa kowane da'awar zuwa bayanai masu ƙarfi ko tushen asali yana hana ƙirƙira ƙirƙira samun hanyar shiga cikin rahoton ko bugawa.
3. Lokacin fassara BLAST ko sakamakon jeri-jeri, wanne ake buƙata don tantance wasan 'aminci'?
- A) Kallon tsayin igiyar kawai
- B) Dogaro kai tsaye ga nau'in sunan da aka bayar da hankali na wucin gadi
- C) Kimanta ma'aunin daidaitawa kamar kashi na ainihi, ɗaukar hoto, da e-daraja tare ✔
- D) Kawai kirga layukan da aka fitar
Bayani: Ma'auni kamar asalin kashi, ɗaukar hoto, da ƙimar e-daraja ana buƙatar ingantacciyar fassarar jeri. Ƙaramar e-darajar tana nuna cewa kamanceceniya ba ta zo daidai ba. Idan ba tare da waɗannan ma'auni ba, fassarar 'wannan sunadaran shine' ba za a iya tabbatar da shi ba kuma yana iya zama abin gani.
4. Me yasa ake amfani da 'daidaitacce p-value' (padj) lokacin gwajin kwayoyin halitta 20,000 lokaci guda a cikin nazarin maganganun bambancin RNA-seq?
- A) Don haɓaka canjin ƙasa
- B) Don sarrafa abubuwan da ba daidai ba saboda gwaje-gwaje da yawa da iyakance ƙimar gano karya ✔
- C) Don rage girman samfurin
- D) Don hanzarta bincike
Bayani: Lokacin da aka gwada dubban kwayoyin halitta a lokaci guda, ɗaruruwan kwayoyin halitta na iya zama 'mahimmanci' ko da kwatsam. Gyaran gwaji da yawa kamar Benjamini-Hochberg yana iyakance ƙimar gano ƙarya. Tare da danyen p-darajar lissafin ya zama kumbura marar kuskure; Amfani da padj yana da mahimmanci don kariyar kimiyya.
5. Menene tarkon da ke faruwa a cikin nazarin omics lokacin da aka sarrafa ƙungiyoyin jiyya guda biyu a ranaku daban-daban, wanda ke haifar da kuskuren bambancin fasaha don bambancin ilimin halitta?
- A) Kuskuren canjin bene
- B) Kodon ingantawa
- C) Tasirin Batch ✔
- D) Tasirin Edge
Bayani: Tasirin tsari shine bambance-bambancen fasaha da ke fitowa daga sarrafa samfuran ta kwanaki daban-daban, na'urori ko mutane kuma shine babban tushen kuskure a cikin binciken omics. Kafin fara bincike, ya zama dole a zana ginshiƙi na PCA kuma bincika ko samfuran sun taru bisa ga yanayin halitta ko tsari.
6. Menene ma'aunin pLDDT ke nufi don hasashen tsarin furotin da aka samar tare da AlphaFold kuma ta yaya ya kamata a yi amfani da shi?
- A) Yana nuna matakin amincewa na samfurin ga kowane yanki; Yakamata a guji iƙirarin da aka dogara akan ƙarancin amincewa ✔
- B) Yana auna yadda furotin ke ninka da sauri kuma ana iya yin watsi da su
- C) Ya tabbatar da cewa furotin yana da madaidaicin tsari wanda aka warware ta hanyar gwaji.
- D) Yana ba da alaƙar docking kai tsaye
Bayani: pLDDT maki ne na amincewa wanda ke nuna yadda kwarin gwiwar ƙirar ke da kowane amino acid. Babban dabi'u (> 90) gabaɗaya abin dogaro ne; ƙananan dabi'u (<50) yawanci suna nuna yankuna marasa tsari ko rashin tabbas. Hanyoyin da'awar da suka dogara da yankuna masu ƙanƙanta na yaudara ne; Dole ne a tabbatar da mahimman tsari ta hanyar gwaji.
7. Ta yaya za a fassara makin docking na ƙwayoyin cuta a ƙirar ƙwayoyi/enzyme?
- A) Ya kamata a yi la'akari da shi azaman cikakkiyar alaƙa.
- B) Yana tabbatar da cewa kwayar halitta ba za ta taɓa ɗaure ba
- C) Kayan aiki ne na dangi don yin odar kwayoyin halitta kafin gwaji; An yanke shawarar ƙarshe ta hanyar auna kusancin gwaji ✔
- D) Shi kaɗai ya isa don amincewar asibiti
Sharhi: Docking maki dangi ne kuma ba sa hasashen ainihin alaƙar ɗaurewa; Adadin tabbataccen ƙarya yana da yawa. Amfanin da ya dace shine a jera dubunnan kwayoyin halitta kafin gwaji da kuma haskaka mafi kyawu. An yanke shawarar ƙarshe ta hanyar auna kusanci na gwaji kamar SPR ko ITC.
8. Menene babban aibi na hanyar 'masu canji ɗaya a lokaci ɗaya' (OFAT) don injiniyan bioprocess da ke neman haɓaka sigogi biyar?
- A) Rashin mu'amala tsakanin sigogi ✔
- B) Koyaushe yana buƙatar ɗan gwaji kaɗan
- C) Yana ƙara ƙarfin ƙididdiga
- D) Yana kawar da tasirin tsari ta atomatik
Bayani: Hanyar OFAT ta rasa hulɗar tsakanin sigogi; watakila babban zafin jiki yana da kyau kawai a ƙananan pH. Zane na gwaje-gwaje (DoE) yana ɗaukar ma'amala kuma yana rage adadin gwaje-gwaje tare da ƙirar ƙira waɗanda ke bambanta sigogi tare da daidaitawa.
9. Menene madaidaicin hanya lokacin da samfurin ingantawa ya ba da shawarar zafin jiki wanda zai kashe al'ada a cikin dakin gwaje-gwaje?
- A) Aiwatar da shawarar kamar yadda yake saboda samfurin ya fi wayo
- B) Ba da aminci da iyakoki na physiological a matsayin ƙuntatawa ga ƙirar da bincika kowace shawara tare da ilimin dakin gwaje-gwaje ✔
- C) Yin watsi da ingantawa gaba ɗaya
- D) Gwada yin watsi da iyakar zafin jiki
Bayani: Samfurin bai san iyakokin ilimin lissafi da aminci ba; mafi girman ilimin lissafi na iya zama haɗari ko ba zai yiwu ba. Hanyar da ta dace ita ce ba da aminci da iyakoki na ilimin lissafi a matsayin ƙuntatawa ga ƙirar kuma duba kowace shawara tare da ilimin ɗakin gwaje-gwaje. Ƙimar jiki tana ƙayyadad da mafi kyawun ilimin lissafi.
10. Wanne ya wajaba a lokacin da ake kimanta sakamakon kirga tantanin halitta mai sarrafa kansa ko rarrabewar haske?
- A) Yarda da sakamakon kai tsaye, saboda samfurin ya fi sauri fiye da mutum
- B) Bayar da rahoton adadin hotuna kawai
- C) Kallon hoton kawai tare da sigina mafi girma
- D) Da hannu tabbatar da fitarwa akan samfurin kuma inganta tare da sarrafawa masu dacewa (ciki har da mara kyau, mara kyau) ✔
Bayani: Dole ne a tabbatar da fitarwa ta atomatik da hannu akan samfurin kuma kowane ma'auni dole ne a inganta shi tare da sarrafawa masu dacewa (tabbatacce, korau, mara kyau, mara kyau). Misali, idan akwai sigina a cikin kulawar da ba ta da kyau, wannan na iya zama autofluorescence; Ba tare da sarrafawa ba, bincike ta atomatik ba zai iya bambanta sigina na gaskiya daga kayan tarihi ba.
11. Menene ya kamata a yi idan shawarar ingantawa a cikin tsarin bioprocess yana ƙaruwa yawan amfanin ƙasa amma ya ɗauki sifa mai inganci (CQA) ba tare da ƙayyadaddun bayanai ba?
- A) Tun da inganci yana da mahimmanci, an fi son zaɓin babban inganci
- B) Ana kiyaye inganci ta hanyar shakatawa da iyakar CQA
- C) An bar yanke shawara ga basirar wucin gadi
- D) Ana ba da fifikon inganci akan inganci; Zaɓin inganci wanda ya faɗi cikin sararin ƙira an fi son ✔
Bayani: A cikin bioprocessing, ingancin trumps yawan amfanin ƙasa; Dole ne a kiyaye iyakokin CQA (tsarki, glycosylation, aiki) don samfurin ya kasance mai aminci da tasiri. Idan ma'anar da ke haɓaka aiki yana lalata inganci, zaɓin ingancin da ya rage a cikin sararin ƙira ya fi dacewa.
12. Menene babban haɗari idan aka gaya wa samfurin harshe don 'nemo labarai 10 akan wannan batu kuma a taƙaita su'?
- A) Samfurin yana gudana a hankali
- B) Samfurin na iya haifar da nassoshi na gaskiya amma babu (ƙirarru) ba tare da yin bincike na gaske ba ✔
- C) Samfurin koyaushe yana samun labarai da yawa
- D) Samfurin yana dawo da tsoffin labarai ne kawai
Bayani: Kayan aikin taɗi na yau da kullun ba sa yin bincike na ainihi; yana samar da amsoshi 'da alama mai yuwuwa' daga ƙwaƙwalwar ajiya. Wannan yana nufin nassoshi na gaskiya amma na karya (mawallafi, mujallu, DOI). Ya kamata a tabbatar da kowane tunani akan ainihin bayanan bayanai (PubMed, DOI) kuma, idan zai yiwu, ya kamata a yi amfani da kayan aikin tushen RAG da ke da alaƙa da ainihin takaddun.
13. Menene madaidaicin hali lokacin da mai amfani ya tambayi AI don maye gurbin da zai ƙara tasirin ƙwayar ƙwayar cuta?
- A) Amsa bukata dalla-dalla saboda ilimi ne
- B) Rage haɗari ta hanyar samar da bayanai kaɗan kawai
- C) ƙin yarda da buƙatar, bayyana cewa yana ƙarƙashin DURC, kuma a ba da rahoto ga tashar samar da kwayoyin halitta ✔
- D) Yi watsi da buƙatar kuma matsa zuwa wani batu
Bayani: Wannan buƙatar buƙatu ce mai cutarwa a ƙarƙashin amfani biyu (DURC). AI yakamata ya ƙi irin waɗannan buƙatun, ya bayyana dalilin da yasa yake haifar da haɗarin amfani biyu, kuma ya ba da rahoton lamarin ga tashar biosafety/da'a na kamfani. Bai kamata a ba da shawarar fasaha da za ta ƙara yuwuwar cutarwa ba.
14. Menene ƙarshe daidai lokacin da rubutun binciken Python da aka rubuta ta hanyar fasaha na wucin gadi yana aiki ba tare da kurakurai ba?
- A) Sakamakon daidai ne saboda lambar tana aiki
- B) Babu buƙatar gwada lambar saboda AI ya rubuta shi
- C) Rashin kurakurai kuma yana ba da garantin sake haihuwa.
- D) Ƙididdiga mai gudana bazai zama daidai ba; Dole ne a tabbatar da fitarwar da ake tsammanin ta sabawa sanannun bayanan gwaji ✔
Bayani: 'Ya yi aiki ba tare da kurakurai' kuma 'ya ba da sakamako daidai' abubuwa biyu ne daban-daban. Lambar na iya dawo da sakamakon da ba daidai ba cikin shiru saboda kuskuren naúrar, ginshiƙi mara kyau, ko matatar da aka rasa. Dole ne a gwada kowace lamba tare da ƙananan bayanan gwaji waɗanda aka san shigarwa da fitarwa; Duk da haka, ya kamata a yi la'akari da abin dogara lokacin da ya ba da sakamakon da ake sa ran.